Skip to main content

Full text of "Sex And Personality Studies In Masculinity And Femininity"

See other formats


McGRAW-HILL PUBLICATIONS IN PSYCHOLOGY 
J. F. DASHIEIXr, PH. D., CONSULTING EDITOR 



SEX AND PERSONALITY 



McGraw-Hill 

Publications in Psychology 
J. F. DASHIELL 

OONBULTINO XDITOR 
Brotfln -PSYCHOLOGY AND THE SOCIAL 

ORDER 

Davit PSYCHOLOGY or LEARNING 
Franz and Gordon PSYCHOLOGY 

Franz and Gordon PSYCHOLOGY 
WORK BOOK 

Gutijord PSYCHOMETRIC METHODS 

Lcwin A DYNAMIC THEORY OF 
PERSONALITY 

Lewin PRINCIPLES OF TOPOLOGICAL 
PSYCHOLOGY 

Maier and Schneirla PRINCIPLES OF 
ANIMAL PSYCHOLOGY 

Mtifessel STUDENT'S GUIDE FOR 
DEMONSTRATIONS OF PSYCHO- 
LOGICAL EXPERIMENTS 

PiUsbury AN ELEMENTARY PSY- 
CHOLOGY OF THE ABNORMAL 

Ruckmick THE PSYCHOLOGY OF 
FEELING AND EMOTION 

Terman and Miles SEX AND PER- 
SONALITY 

Wattin PERSONALITY MALADJUST- 
MENTS AND MENTAL HYGIENE 



SEX AND PERSONALITY 

Studies in 
^Masculinity and Femininity 



BY 
LEWIS M. TERMAN AND CATHARINE COX MILES 

Assisted by 

JACK W. DUNLAP E. ALICE MCANULTY 

HAROLD K. EDGERTON QUINN MCNEMAR 

E. LOWELL KELLY MAUD A. MERRILL 

ALBERT D. KURTZ FLOYD L. RUCH 

HORACE G. WYATT 



FIRST EDITION 



McGRAW-HILL BOOK COMPANY, INC 

NEW YORK AND LONDON 
1936 



COPYRIGHT, 1936, BY THE 
MCGRAW-HILL BOOK COMPANY, INC. 

PRINTED IN THE UNITED STATES OF AMERICA 

All rights reserved. This book, or 

parts thereof, may not be reproduced 

in any form without permission of 

the publishers. 



THE MAPLE PRESS COMPANY, YORK, PA. 



PREFACE 

Sex differences in personality and temperament are matters of 
universal human interest. Among all classes of people, from the 
most ignorant to the most cultivated, they provide an inexhausti- 
ble theme for light conversation and humorous comment. They 
have always been and perhaps always will be one of the chief 
concerns of novelists, dramatists, and poets. They are rapidly 
coming to be recognized as one of the central problems in anthro- 
pology, sociology, and psychology. 

It is well that they should be so recognized; for sex differences 
are more than a perennial stimulus to idle speculation, wit, and 
literary art. Mass theories in regard to them are one of the most 
potent of all the forces that operate in the shaping of human 
societies, from the most primitive to the most modern. In 
every culture they help to determine the accepted patterns of 
family life, of education, of industry, and of political organization. 
Anthropologists have shown that the standard patterns of male 
and female behavior present every shade and degree of variety, 
even to the well-nigh complete reversal of the roles commonly 
prevalent in Occidental society. Sometimes the male and female 
patterns are virtually nonoverlapping and arbitrarily enforced; 
sometimes they are less differentiated and less rigidly maintained. 
Nearly always, however, there is a recognized dichotomy which 
seems to be based on the tacit assumption that men and women, 
by the mere fact of their sex, differ more than they resemble, 
and that the members of either sex considered alone make up a 
population which, biologically and psychologically, is relatively 
homogeneous. 

Many students of human nature, especially the anthropologist, 
the psychiatrist, and the psychologist, have questioned this 
assumption. The anthropologist encounters so many varieties 
of dichotomy with respect to masculine and feminine behavior 
that it seems impossible to explain them wholly in terms of 



vi PREFACE 

biological factors. The psychiatrist in his clinical practice finds 
a large proportion of his patients among men and women who, 
because of either their exceptional nature or their exceptional 
nurture, have had major difficulties in adjusting to the sexual 
patterns society has assigned to them. The psychologist of 
individual differences has so often found his results in opposition 
to popular views with respect to the existence of human types; 
has so often discovered wide variation within and consequent 
overlapping of alleged types; has become so familiar with the 
possibilities of psychological conditioning to gradations of 
behavior, that he, too, views with suspicion the categorical 
explanation of any aspect of human nature in terms of well- 
defined dichotomies. 

Unfortunately, investigations of masculinity and femininity 
have been retarded by lack of definiteness with respect to what 
these terms should connote. Gross departures from even a vaguely 
defined norm have of course long been recognized, but in the 
absence of quantitative methods the less extreme deviations are 
overlooked or misunderstood. The present situation resembles 
that which obtained a few decades ago with respect to mental 
deficiency or insanity, when, in default of quantitative concepts, 
the psychiatrist classified his subjects as "normal" or "feeble- 
minded," "normal" or "insane," etc. Thanks largely to Binet 
and his successors on the one hand, and to modern psychiatry 
on the other, no competent investigator in abnormal psychology 
now regards such classifications as adequate or even possible. 
No more adequate, we believe, is the classification of subjects as 
"normal" or "invert" with respect to masculinity or femininity. 

The purpose of the investigations here reported has been the 
accomplishment in the field of masculinity-femininity of something 
similar to Binet's early achievement in the field of intelligence 
a quantification of procedures and of concepts. No one can better 
realize than the authors how imperfectly they have succeeded in 
their pioneer attempt. The problem of temperament is no less 
complex than the problem of abilities. The concepts of mascu- 
linity and femininity are even more vague than the nineteenth 
century concepts of intelligence. Clarity and exactness are 
seldom attained by a single effort. Indeed, the investigations 



PREFACE vii 

to be reported have been shaped by the conviction that only tte 
simplest and roughest kind of quantification is at present pos- 
sible, and that any attempt at exact measurement of the traits 
in question would, in the present state of psychometric develop- 
ment, be fatuous and unprofitable. The experiment will have 
justified itself if in some degree it opens the way to more precise 
measurements and methods. 

During the eleven years since the experiment was undertaken, 
the authors have accumulated a great load of indebtedness to 
those who have helped to bring the study to its present stage of 
development. Most of all, they are under obligation to the 
Committee for Research on Problems of Sex, of the National 
Research Council, for a series of grants which made the investiga- 
tion possible, and to Mr. George Frick for additional financial 
support. The names of the research assistants who have con- 
tributed most heavily appear on the title page, but the authors 
wish to record here their appreciation of the skilled and devoted 
services rendered by these younger colleagues. Drs. Jack W. 
Dunlap, Harold K. Edgerton, Albert D. Kurtz, Floyd L. Ruch, 
and Miss E. Alice McAnulty assisted in the technical and statisti- 
cal detail of the test standardization. Dr. Quinn McNemar 
collaborated in this part of the work, in the standardization of 
Exercise III as an intelligence test, and in a preliminary study of 
the M-F scores of married couples. Special recognition is due 
to Dr. E. Lowell Kelly for providing the data presented in the 
three chapters on homosexuality in males, to Professor Maud A. 
Merrill for providing the case information on which Chapter XIV 
is based, and to Mr. Horace G. Wyatt for bringing together the 
detailed material for the chapter on "Sex Temperaments as 
Revealed by the M-F Test." 

The authors gratefully record other valued assistance. Mary 
A. Bell contributed in the development of the ink-blot association 
technique and administered other tests to various populations. 
Dr. Psyche Cattell helped assemble the test items in a first trial 
series, particularly those for the interest test. Madeline Frick 
carried out the study of trait ratings of college students described 
in Chapter V. The teachers' ratings of masculinity and femininity 
mentioned in the same chapter were provided by Lela Gillen. 



Viii PREFACE 

Helen McCreery assisted in bringing together some of the data 
for the case notes of Chapter XV. Much of the computational 
work was done by Kirk Miles. For assistance in proofreading 
the authors are indebted to Gladys Buttenwieser, Helen 
McCreery, and Frances S. Shaffer. Tests were administered 
to the students at the University of California by Professor 
George M. Stratton, at the University of Oregon by Pro- 
fessor Howard Taylor, and at the University of Utah by Professor 
Dorothy Nyswander. Mention should also be made of those 
who aided the research by securing data from other school and 
college groups. These include especially Professors Charles A. 
Breck, Jack W. Dunlap, LI. Wynn- Jones, EUen B. Sullivan, 
J. St. Clair Price, Drs. Esther A. Gaw and Hazel M. Stanton, Miss 
Rosalind Cassidy, Mr. Walter R. Chivers, and Miss Helen 
Marshall. Finally, the authors acknowledge with appreciation 
their indebtedness to the thousands of subjects who have pro- 
vided the fundamental data and to the teachers and school 
officials who made possible the administration of the test to 
school populations. 

The senior author is responsible for the original plan of the 
investigation, for bringing the testing techniques to essentially 
their present form, and for the plan and preparation of the 
present volume. The junior author assisted in the final stand- 
ardization of the two forms of the test, arranged for administra- 
tion of the tests to all the main populations with the exception 
of the male homosexuals, directed jointly with the senior author 
the work of most of the assistants for four years, and provided the 
original draft of Chapters V to X, and the material in Appendix 
VII. The authors have collaborated in the final revision of the 
manuscript. 

LEWIS M. TERMAN. 

CATHARINE Cox MILES. 

August, 1936. 



CONTENTS 

PAGE 

PREFACE v 

LIST OF ILLUSTRATIONS ri 

CHAPTER I 

RATIONALE OF THE MASCULINITY-FEMININITY TEST .... 1 

CHAPTER II 

ORIGIN OF THE M-F TEST 11 

CHAPTER III 

EXTENSION OF THE M-F TEST TO NEW TECHNIQUES 18 

CHAPTER IV 

THE M-F TEST AS A WHOLE 52 

CHAPTER V 

CORRELATIONS OF M-F SCORE WITH PHYSICAL MEASUREMENTS 
AND TRAIT RATINGS 80 

CHAPTER VI 

CORRELATION OF M-F SCORE WITH PERSONALITY AND ACHIEVE- 
MENT MEASURES 98 

CHAPTER VII 

RELATION OF M-F SCORE TO AGE, EDUCATION, AND INTELLI- 
GENCE 122 

CHAPTER VIII 
RELATION OF M-F SCORE TO OCCUPATION 157 

CHAPTER DC 

EFFECT OF INTERESTS ON M-F SCORE 195 

CHAPTER X 

RELATION OF M-F SCORE TO THE DOMESTIC MILIEU 223 

CHAPTER XI 

A STUDY OF MALE HOMOSEXUALS 239 

iz 



x CONTENTS 

PAGE 
CHAPTER XII 

A TENTATIVE SCALE FOR THE MEASUREMENT OF SEXUAL 
INVERSION IN MALES 259 

CHAPTER XIH 
CASE STUDIES OF HOMOSEXUAL MALES 284 

CHAPTER XIV 

CASE STUDIES: MASCULINE AND FEMININE TYPES or DELIN- 
QUENT GIRLS 321 

CHAPTER XV 
CASE NOTES: MISCELLANEOUS 342 

CHAPTER XVI 
SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 371 

CHAPTER XVII 

INTERPRETATIONS AND CONCLUSIONS 451 

APPENDIX I 
NORMS FOR TOTAL SCORE 471 

APPENDIX II 

NORMS FOR EXERCISES 1 TO 7 474 

APPENDIX III 

CONVERSION TABLES 480 

APPENDIX IV 
FORMS A AND B OF THE M-F TEST, WITH SCORING KEYS . . . 482 

APPENDIX V 

TRAIT RATINGS 555 

APPENDIX VI 
CASE HISTORY RECORD FOR MALE HOMOSEXUALS 564 

APPENDIX VII 
GROUP PROFILE COMPARISONS 570 

INDEX 589 



LIST OF ILLUSTRATIONS 

FIGURE PAGE 

1 . ILLUSTRATION OF THE MEASURE OF OVERLAPPING ... . ... 67 

2. M-F SCORES OF MEN IN RELATION TO INTELLIGENCE AND SCHOLARSHIP 109 

3. M-F SCORES OF WOMEN IN RELATION TO INTELLIGENCE AND SCHOLAR- 

SHIP 113 

4. M-F SCORES OF ATHLETIC AND OTHER COLLEGE-WOMEN'S GROUPS. . 116 

5. AGE TRENDS OF MEAN M-F SCORE . . 123 

6. AGE TRENDS OF MEAN M-F SCORE IN RELATION TO SCHOOLING. ... 139 

7. AGE INFLUENCE UPON MEAN M-F SCORES OF CERTAIN MALE GROUPS 151 

8. AGE INFLUENCE UPON MEAN M-F SCORES OF CERTAIN FEMALE GROUPS 152 

9. MEAN M-F SCORES OF OCCUPATIONAL AND OTHER GROUPS (MALES). . 160 

10. MEAN M-F SCORES OF OCCUPATIONAL AND OTHER GROUPS (FEMALE) . 181 

11. MEAN M-F SCORES OF ADULT MALES WITH MUCH INTEREST OR LITTLE 

OR No INTEREST IN VARIOUS FIELDS .... 205 

12. MEAN M-F SCORES OF ADULT FEMALES WITH MUCH INTEREST OR LITTLE 

OR No INTEREST IN VARIOUS FIELDS 213 

13. PERCENTAGES OF MEN AND WOMEN PROFESSING MUCH INTEREST OR 

LITTLE OR No INTEREST IN VARIOUS FIELDS . . 221 

14. DISTRD3UTION OF M-F SCORES OF 77 PASSIVE MALE HOMOSEXUALS, 

98 HIGH-SCHOOL BOYS, AND 90 HIGH-SCHOOL GIRLS 241 

15. DISTRD3UTION OF M-F SCORES OF 77 PASSIVE MALE HOMOSEXUALS, 

46 ACTIVE MALE HOMOSEXUALS, AND 42 REGULAR ARMY MEN . . . 242 

16. DISTRIBUTIONS OF "I" SCORES OF 82 PASSIVE MALE HOMOSEXUALS AND 

46 NORMAL MALES WITH Low M-F SCORES 263 



^Acknowledgment 



THIS VOLUME OF STUDIES 

WAS MADE POSSIBLE BY A SERIES OF GRANTS FROM 

THE COMMITTEE FOR RESEARCH ON PROBLEMS OF SEX 

OF THE NATIONAL RESEARCH COUNCIL 



SEX AND PERSONALITY 

CHAPTER I 

RATIONALE OF THE MASCULINITY-FEMININITY TEST 

The belief is all but universal that men and women as contrast- 
ing groups display characteristic sex differences in their behavior, 
and that these differences are so deep seated and pervasive as to 
lend distinctive character to the entire personality. That 
masculine and feminine types are a reality in all our highly 
developed cultures can hardly be questioned, although there is 
much difference of opinion as to the differentiae which mark them 
off and as to the extent to which overlapping of types occurs. It 
is true that many social trends, of which the recent development 
of psychological science is one, have operated to reduce in the 
minds of most people the differences that had long been supposed 
to separate the sexes in personality and temperament. Intelli- 
gence tests, for example, have demonstrated for all time the 
falsity of the once widely prevalent belief that women as a dass 
are appreciably or at all inferior to men in the major aspects of 
intellect. The essential equality of the sexes has further been 
shown by psychometric methods to obtain also in various special 
fields, such as musical ability, artistic ability, mathematical 
ability, and even mechanical ability. The enfranchisement of 
women and their invasion of political, commercial, and other 
fields of action formerly reserved to men have afforded increas- 
ingly convincing evidence that sex differences in practical abilities 
are also either nonexistent or far less in magnitude than they have 
commonly been thought to be. 

But if there is a growing tendency to concede equality or near 
equality with respect to general intelligence and the majority of 

l 



2 SEX AND PERSONALITY 

special talents, the belief remains that the sexes differ funda- 
mentally in their instinctive and emotional equipment and in the 
sentiments, interests, attitudes, and modes of behavior which 
are the derivatives of such equipment. It will be recognized 
that these are important factors in shaping what is known as 
personality, hence the general acceptance of the dichotomy 
between the masculine and feminine personality types. The 
belief in the actuality of M-F types remains unshaken by the fact, 
abundantly attested, that observers do not agree in regard to the 
multitudinous attributes which are supposed to differentiate 
them. Although practically every attribute alleged to be char- 
acteristic of a given sex has been questioned, yet the composite 
pictures yielded by majority opinion stand out with considerable 
clearness. 

In modern Occidental cultures, at least, the typical womanjs 
believed to differ from the typical man in the greater richness and 
variety of her emotional life and in the extent to which her every- 
day behavior is emotionally determined. In particular, she is 
believed to experience in greater degree than the average man the 
tender emotions, including sympathy, pity, and parental love; to 
be more given to cherishing and protective behavior of all 
kinds. Compared with man she is more timid and more readily 
overcome by fear. She is more religious and at the same time 
more prone to jealousy, suspicion, and injured feelings. Sexually 
she is by nature less promiscuous than man, is coy rather than 
aggressive, and her sexual feelings are less specifically localized in 
her body. Submissiveness, docility, inferior steadfastness of 
purpose, and a general lack of aggressiveness reflect her weaker 
conative tendencies. Her moral life is shaped less by principles 
than by personal relationships, but thanks to her lack of adven- 
turousness she is much less subject than man to most types of 
criminal behavior. Her sentiments are more complex than man's 
and dispose her personality to refinement, gentility, and pre- 
occupation with the artistic and cultural. 

But along with the acceptance of M-F types of the sort we have 
sketchily delineated, there is an explicit recognition of the exist- 
ence of individual variants from type : the effeminate man and the 



RATIONALE OF THE MASCULINITY-FEMININITY TEST 3 

masculine woman. Grades of deviates are recognized ranging 
from the slightly variant to the genuine invert who is capable of 
romantic attachment only to members of his or her own sex, 
although, as we shall see later, judges rating their acquaintances 
on degree of masculinity and femininity of personality seldom show 
very dose agreement. Lack of agreement in such ratings proba- 
bly arises from two sources : (1) varying opinion as to what factors 
truly differentiate the M-F types, and (2) varying interpretations 
of specific kinds of observed behavior. Some raters base their 
estimates largely upon external factors, such as body build, 
features, voice, dress, and mannerisms; others, more penetrating, 
give larger weight to the subtle aspects of personality which 
manifest themselves in interests, attitudes, and thought trends. 

For many reasons, both practical and theoretical, it is highly 
desirable that our concepts of the M-F types existing in our pres- 
ent culture be made more definite and be given a more factual 
basis. Alleged differences between the sexes must give place to 
experimentally established differences. A measure is needed 
which can be applied to the individual and scored so as to locate 
the subject, with a fair degree of approximation, in terms of 
deviation from the mean of either sex. Range and overlap of the 
sexes must be more accurately determined than is possible by 
observational and clinical methods. 

It is to this end that the researches set forth in the present 
volume have been directed. As the result of empirical investiga- 
tions, over a period of several years, a test of mental masculinity 
and femininity has been devised which is based upon actual differ- 
ences between male and female groups ranging in age from early 
adolescence to life's extreme. The reliability of the instrument, 
in the sense of consistency of its verdicts, qualifies it for the com- 
parative study both of groups and of individuals. The test has 
been applied to samplings of populations of each sex differing in 
age, education, occupation, cultural interests, and familial back- 
ground. Two groups of male and a few female homosexuals have 
been studied. Scores on the M-F test have been correlated with 
physical measurements and with rated estimates of several person- 
ality variables. The possibility of subjects faking their scores 



4 SEX AND PERSONALITY 

has been investigated. The mean, the range, and the overlap- 
ping of sex groups have been determined for numerous populations. 
The degree to which M-F traits enter into marital selection has 
been measured, and (by the use of a modified M-F test) spouse 
resemblance in M-F score has been correlated with marital 
happiness. The test items carrying masculine weights have been 
compared with those carrying feminine weights in an effort to 
arrive at a more exact and meaningful description of character- 
istic sex differences as they exist in the culture of our time and 
country. 

The M-F test is made up in two equivalent forms, A and B . It 
is entirely a pencil-and-paper test, of the questionnaire variety, 
composed of 910 items (Form A, 456; Form B, 454). In the 
interests of anonymity, as well as to insure speed in administration 
and scoring, all the items are responded to by checking one of 
four, three, or two multiple responses. The test is administered 
without a time limit, a single form requiring forty to fifty minutes 
for the majority of subjects, rarely an hour. It is hardly appli- 
cable to subjects of less than seventh-grade education and ability. 
The scoring is done by the use of stencils and is therefore entirely 
objective. Scoring by the Hollerith machine is possible but is not 
economical except for fairly large populations. Each response 
carries a weight of one and is scored either + or , that is, 
masculine or feminine. Weighted scores were originally used but 
were discarded in favor of unit weights for reasons which will be 
given in Chapter IV. 

N \The purpose of the test has been disguised by the title, "Atti- 
tude-Interest Analysis." This was made necessary by the fact 
that subjects who know what the test is intended to measure are 
able to influence their scores so greatly as to invalidate them 
entirely. 1 

Both forms of the test are reproduced in Appendix I, with + 
and signs inserted to show how each item is scored. It will be 
noted that in subtests 2, 3, 5, and 6 omissions as well as actual 
responses are scored. There are seven subtests (or "exercises") 
as follows: 

1 For data on this point see pp. 77 Jf. 



RATIONALE OF THE MASCULINITY-FEMININITY TEST 





- 


Form A 


Form B 


Exercise 1. 
Exercise 2. 
Exercise 3. 


Word Association.. . . . . 
Ink-blot Association. .. . 
Information 


60 items 
18 items 
70 items 


60 items 
18 items 
70 items 


Exercise 4. 
Exercise 5. 


Emotional and Ethical Response 
Interests 


105 items 
119 items 


105 items 
118 items 


Exercise 6. 
Exercise 7. 


Personalities and Opinions ... . 
Introvertive Response.. 


42 items 
42 items 


41 items 
42 items 


Total.. . 




456 items 


454 items 











Examination of the seven exercises will show that with the excep- 
tion of No. 6 each is composed of items of a fairly homogeneous 
type. Exercise 6 contains two parts, the first of which logically 
belongs with Exercise 5. The make-up of this exercise was influ- 
enced by space requirements in printing the blanks, and the same 
fact accounts for the addition to Exercise 4 of a small group of 
items which logically belong with Exercise 6. This does not 
greatly matter, as it was not the intention that each of the seven 
exercises should necessarily measure a unitary trait. All of them 
together present a wide sampling of sex differences and it is the 
total score with which we are chiefly concerned. The range of 
scores in the general population of adults is roughly as follows: for 
males, from +200 to 100, with a mean of +52 and S.D. of 50; 
for females, from +100 to -200, with a mean of -70 and S.D. of 
47. The score, as we shall see, is influenced by age, intelligence, 
education, interests, and social background, and to such an extent 
that groups differing in these respects often show markedly con- 
trasting score distributions. 

The M-F test does not represent any radical departure in 
principle from methods previously employed in the study of sex 
differences. A majority of the earlier studies also resulted in 
comparisons stated in quantitative terms. For example, the 
sexes were found to differ numerically in the number liking or 
playing particular games ; the number liking or disliking particular 
colors, books, or school subjects; the number preferring this or 
that occupation, literary style, historical character, or ideal; or 
they were found to differ in the degree of introversion, dominance, 



6 SEX AND PERSONALITY 

inferiority feeling, conservatism, emotional stability, sense of 
humor, religious attitudes, or other alleged traits. The gresent 
^study differs from its predecessors primarily in the fact that it 
represents a more systematic attempt to sample sex differences in 
a large variety of fields in which such differences are empirically 
demonstrable. The literature of the subject, both theoretical 
and experimental, was canvassed for dues as to the types of test 
responses most likely to reveal sex differences. Of the thousands 
of items which have been tried out only those have been retained 
which satisfied in some degree this criterion of discriminative 
capacity. The test is based, not upon some theory as to how the 
sexes may differ, but upon experimental findings as to how they do 
differ, at least in the present historical period of the Occidental 
culture of our own country. 

Thejpurpose of the M-F test is to enable the clinician or other 
investigator to obtain a more exact and meaningful, as well as a 
more objective, rating of those aspects of personality in which the 
sexes tend to differ. More specifically, the purpose is to make 
possible a quantitative estimation of the amount and direction 
of a subject's deviation from the mean of his or her sex, and to 
permit quantitative comparisons of groups differing in age, 
intelligence, education, interests, occupation, and cultural milieu. 
By comparing the responses made by groups of subjects on the 
several parts of the test we secure a basis for qualitative as well as 
quantitative studies of sex differences. 

The M-F test rests upon no assumption with reference to the 
causes operative in determining an individual's score. These 
may be either physiological and biochemical, or psychological 
and cultural; or they may be the combined result of both types of 
influence. The gim has been merely to devise a test which would 
measure existing differences in mental masculinity and feminity, 
however caused. It is only when a test of this kind has been 
made available that it becomes possible to investigate with any 
degree of precision the influence of the numerous physical, social, 
and psychological factors that may affect a subject's rating. 

At present extreme differences of opinion are to be found with 
reference to this question. Many if not the majority of students 



RATIONALE OF THE MASCULINITY-FEMININITY TEST 7 

of homosexuality accept the theory that sexual inversion is not a 
product of psychological conditioning but is inborn. This seems 
to be also the almost universal belief of homosexuals themselves, 
though one would of course hesitate to give much weight to their 
unsupported opinion. So far as animals below man are con- 
cerned, the facts available point unmistakably to the conclusion 
that maleness and femaleness of behavior are biochemically 
determined. Transplantation of male gonadal tissue info the 
ovariectomized domestic fowl induces not only male secondary 
sexual characters but also typical male behavior. A correspond- 
ing effect in the reverse direction is produced by ovarian trans- 
plants in the castrated male fowl. Spontaneous sex reversals in 
pigeons and domestic fowls, ordinarily initiated by pathological 
conditions in the gonads, have been described in detail by 
Riddle, Crew, and others. In one case, reported by Crew, a hen 
which had been the mother of two broods of chicks lost her ovaries 
from a tubercular infection, developed male sexual organs, and 
later became the father of another brood of offspring. 1 The 
injection of the female hormone into the blood of virgin female 
white rats is followed by maternal behavior patterns otherwise 
observed only in late pregnancy and the postdelivery period, 
particularly nest building and retrieving of young. The impres- 
sive contrast between the "personality" of stallion and gelding, or 
of bull and steer, is familiar to every farm boy. In a majority of 
mammals and birds in which castrated subjects have been 
experimentally studied, the effect of this operation on the male 
is to produce a temperament (personality) more or less approxi- 
mating that of the normal female. Castration of the female, on 
the other hand, has little effect except upon the specific patterns 
concerned with mating and with maternal behavior. 

Strong as the temptation is to draw inferences from lower 
mammals and fowls to man, it must be resisted in view of man's 
enormously greater modifiability by psychological conditioning. 
Although early castration of the human male is well known to be 
followed by the development of a temperament lacking a number 

1 CREW, F. A. E., Abnormal sexuality in animals. III. Sex reversal. Quart. 
Bid., 1927, 2, 427-441. 



8 SEX AND PERSONALITY 

of the usual attributes of masculinity, we are not able to dis- 
entangle the biochemical from the psychological factors which 
may have combined to produce the total result. 

Our investigations offer considerable evidence of the influence 
of nurture on the masculinity and femininity of human personality, 
although we regard our results as far from conclusive. Perhaps 
cross-parent fixation followed by homosexuality is one of the most 
convincing illustrations of environmental influence, but even here 
biochemical abnormalities have not been ruled out as possible 
contributing causes of the mother-son or father-daughter attrac- 
tion. In view of the scanty and conflicting data of strictly 
scientific nature it would seem that the only just course is to keep 
an open mind and to investigate every possible influence that 
is subject to even approximate measurement. One awaits with 
special interest the results of biochemical studies of homosexual 
subjects, particularly those which will establish the behavioral 
effects induced by synthetic preparations of male and female 
hormones. In these and other experiments it is hoped that the 
M-F test will prove a helpful research tool. 

In view of the myriad known physiological and biochemical 
differences between men and women, any degree of overlap of the 
sexes on a masculinity-femininity test of the type used in the 
investigations to be described must be regarded as psychologically 
and sociologically very significant. If it can be shown that 
despite the biological dichotomy between males and females of 
the genus Homo a few members of each sex rate as masculine or as 
feminine as the average member of the opposite sex, a heavy 
burden of proof devolves upon anyone who doubts the weighty 
influence of environment in shaping the patterns of male and 
female behavior. Certainly any such finding must constitute a 
challenge of the first order to the search for possible physiological 
and biochemical correlates of extreme deviation from the respec- 
tive sex norms in M-F score. It is only by the discovery of such 
correlates that it will become possible to establish any definite 
limits to the effect of environmental influences. In the mean- 
time, studies of sex differences by the use of subjective methods 
will remain of indeterminate value. 



RATIONALE OF THE MASCULINITY-FEMININITY TEST 9 

At this point it may be well to give a few words of warning in 
regard to possible misuses of the M-F test. 

In the first place, one must bear in mind that since the test does 
not sample every conceivable kind of sex difference and does not 
sample with perfect reliability even those fields which it attempts 
to cover, the score it yields cannot be regarded as an adequate 
index of the totality of a subject's mental masculinity and femi- 
ninity. It would doubtless be possible to find enough additional 
valid items to make up several other tests as lengthy as the one 
we have devised. This would especially be the case if the test 
were not exclusively of the pencil-and-paper variety. 

Secondly, a more serious limitation to the present usefulness of 
the test lies in the fact that as yet too little is known about the 
behavior correlated with high and low scores. Painstaking 
clinical studies of large numbers of high-scoring and low-scoring 
subjects will be necessary to remedy this defect. Most emphatic 
warning is necessary against the assumption that an extremely 
feminine score for males or an extremely masculine score for 
females can serve as an adequate basis for the diagnosis of homo- 
sexuality, either overt or latent. It is true, as we shall show, that 
male homosexuals of the passive type as a rule earn markedly 
feminine scores, and that the small number of female homo- 
sexuals of the active type whom we have tested earned high 
masculine scores. That the converse of these rules is in accord 
with the facts, we have no assurance whatever; indeed, our 
findings indicate that probably a majority of subjects who test as 
variates in the direction of the opposite sex are capable of making 
a perfectly normal heterosexual adjustment. Mental masculinity 
or femininity is at most only one of a number of factors predis- 
posing to homosexuality; one must even bear in mind the possi- 
bility that it may be a secondary rather than a primary condition, 
an effect rather than a cause. 

Used with suitable precaution we believe that the M-F test will 
be found valuable both as a clinical and as an investigational tool. 
The application of a single form is adequate for comparison of 
population groups and also for securing approximate ratings of 
individual subjects. When it is important that a subject's rating 



10 SEX AND PERSONALITY 

be as accurate as possible, both forms of the test should be admin- 
istered and the average taken of the two scores. Types of 
investigation in which the test should be helpful include, among 
others, the relationship of masculinity and femininity of tempera- 
ment to body build, metabolic and other physiological factors, 
excess or deficiency of gonadal and other hormone stimulation, 
and homosexual behavior, and to such environmental influences 
as parent-child attachments, number and sex of siblings, sex of 
teachers, type of education, marital compatibility, and choice 
of friends or of occupations. It will be especially interesting to 
compare M-F differences in different cultures, and in the same 
culture at intervals of one or more generations. In short, the 
measurement of M-F differences will make it possible greatly to 
expand our knowledge of the causes which produce them. 

Our primary task has been to throw light on the meaning of the 
M-F score. For this reason we have presented only a relatively 
brief summary of the extensive experimental work which was 
necessary to bring the test to its present form, and have devoted 
the bulk of our volume to the relationships found to obtain 
between M-F score and other variables, including physique, 
personality traits as rated and measured, achievement, age, edu- 
cation, intelligence, occupational classification, interests, domestic 
milieu, delinquency, and homosexuality. It is hoped that the 
reader will thus be sufficiently impressed by the multiplicity of 
factors which go to determine an individual's score and by the 
complexity of interaction among them. In the interest of con- 
creteness and in order to illustrate some of the major questions 
that arise in score interpretation, three chapters have been 
devoted to case studies. We believe that many of these will be of 
surpassing interest to the general reader as well as to the profes- 
sional student of personality and temperament 



CHAPTER II 
ORIGIN OF THE M-F TEST 

The idea of developing a test of masculinity and femininity first 
occurred to the senior author in 1922 in connection with an 
investigation of intellectually superior children. One division of 
that investigation had for its purpose comparison of gifted and 
unselected children with respect to their interest in, practice of, 
and knowledge about plays, games, and amusements. Each of 
90 such activities was rated three times by each subject; first for 
acquaintance with it, secondly for interest in it, and thirdly for 
frequency"with which it was practiced. There followed a list of 
45 questions about experience or accomplishment in a wide 
variety of activities hardly to be classed as plays or games, such 
as " Have you ever cooked a meal ?" " Havfe you ever taken part 
in a play?" etc. Finally, there were 123 information questions 
designed to test the subject's actual knowledge about the plays, 
games, and other activities. (Examples: "A singing game is 
follow-the-leader, London Bridge, poison"; "A game in which 
you must not smile is fruit-basket, old-witch, tin-tin.") 

The test was given to 303 boys and 251 girls of IQ 140 or above, 
ages 6 to 14, and to a control group of 225 unselected boys and 249 
unselected girls, ages 8 to 17. 1 

Sex differences between unselected boys and girls were com- 
puted for the composite of the two ratings given by the subjects 
for interest in and practice of the 90 plays, games, and activities. 
Next a masculinity index was computed for each of the activities, 

1 Subjects in the control group could not read well enough to take the test 
below the age of 8 years. As the test was given only through the eighth grade, 
in which gifted children above the age of 14 are rarely found, the age range for 
the gifted group does not run as high as for the control group. The unselected 
children accordingly have a considerable advantage in the comparison of the 
entire gifted and control group; only for ages 8 to 14 inclusive are the two groups 
strictly comparable. 

11 



12 



SEX AND PERSONALITY 



based upon the sex differences found in the control group. These 
indices ranged from 2 to 24, those above 13 indicating greater 
preference by boys, those below 13 greater preference by girls. 
They are shown in Table 1. 

TABLE 1. MASCULINITY AND FEMININITY INDICES OF NINETY ACTIVITIES 



13 Anty over 


18 Fish 


12 Post office 


12 Authors 


13 Follow the leader 


1 1 Puss in corner 


14 Backgammon 


20 Football 


15 Racing or jumping 


19 Baseball 


13 Fox and geese 


13 Red rover 


16 Basketball 


12 Fox and hounds 


9 Ring around rosy 


20 Bicycle 


16 Garden work 


13 Roly-poly 


14 Billiards 


13 Geography cards 


15 Row a boat 


12 Blackinan 


9 Guessing games 


6 School, play 


10 Blindfold 


14 Handball 


8 Sewing 


17 Bow and arrow 


1 1 Hide-and-seek 


15 Shinny 


14 Bowling 


14 Hike 


21 Shoot 


20 Boxing 


13 History cards 


1 1 Simon says thumbs up 


13 Cards 


15 Hoops 


10 Skate 


9 Cat and mouse 


4 Hopscotch 


17 Ski 


9 Charades 


14 Horseback riding 


13 Snap 


14 Checkers 


5 House, play 


15 Snap the whip 


14 Chess 


IS Hunt 


17 Soccer 


11 Church, play 


9 In and out window 


1 1 Solve puzzles 


14 Coast or toboggan 


10 Jacks tones 


16 Stilts 


5 Cook a meal 


12 Jacks traws 


8 Store, play 


13 Crokinole 


9 Jump rope 


15 Swim 


13 Croquet 


20 Kites 


11 Tag 


8 Dance 


7 Knit or crochet 


12 Tennis 


11 Dare base 


14 Leapfrog 


13 Tiddly-winks 


2 Dolls, play 


9 London Bridge 


24 Tools, use 


13 Dominoes 


19 Machinery, work with 


19 Tops 


3 "Dress up," play 


20 Marbles 


17 Tug of war 


10 Drop handkerchief 


13 Parchesi 


13 Volleyball 


14 Duck on rock 


13 Pom-pom pull-away 


13 Word building 


9 Farmer in the dell 


16 Pool 


20 Wrestling 



By use of the above data it was possible to derive a masculinity 
rating of each subject based upon the score weights given to 
"masculine" and "feminine" activities. The distributions of 
these masculinity ratings are given in Table 2 separately by sex, 
for the gifted and control groups. 

A certain amount of historic interest attaches to the data pre- 
sented in Table 2. Here, for the first time, we have distributions 



ORIGIN OF THE Jf-F TEST 13 

TABLE 2. MASCULINITY RATINGS OP GiFTED AND UNSELECTED CHILDREN* 





Control group 


Gifted group 


Boys 


Girls 


Boys 


Girli 


20 


1 




1 




19 


1 




3 




18 


4 




6 




17 


14 


1 


33 


1 


16 


34 




76 


1 


IS 


40 


1 


92 


6 


14 


34 


7 


62 


16 


13 


31 


17 


8 


23 


12 


1 


56 


6 


54 


11 




39 


2 


69 


10 


1 


36 




34 


9 




15 




20 


8 




5 




4 


7 








4 


6 










5 






1 




N 


161 


177 


290 


232 


Mean 


14.90 


11.23 


15.22 


11.35 


S.D. 


1 49 


1 45 


1 45 


1.65 



* Compiled from Tables 153c and 1536, Genetic studies of genius, vol. I., pp. 411-412, by 
Lewis M. Tennan, et a/., Stanford Univ. Press, 1925. 

by sex of masculinity ratings based upon an extensive sampling of 
behavior responses. We see that the range for each sex is very 
wide and that there is a considerable amount of overlap between 
the sexes; that there are indeed a few members of each sex who 
test beyond the mean of the opposite sex. Incidentally, though 
the fact is not of prime interest for our present purpose, little 
difference in masculinity was found for the gifted and control 
groups at any age. 

It is possible that the masculinity test would have been allowed 
to rest at this point but for facts which came to light in a com- 
parison of scores and case-history data for a number of subjects 
deviating greatly from their sex norm. Some of these compari- 
sons indicated that the scores tended to be correlated with general 
masculinity and femininity of behavior and to reveal an impor- 
tant line of cleavage in personality and temperament. One of the 



14 SEX AND PERSONALITY 

deviating cases in particular furnished considerable motivation to 
further experimentation in the field of M-F testing. This was the 
gifted boy who can be identified in Table 2 as receiving a 
masculinity rating not only below all the other boys, of either the 
gifted or the control groups, but also below that of any girl. The 
following account of the case offers spectacular evidence of 
the significance that may attach to M-F ratings of the kind in 
question. 

An assistant who was tabulating the masculinity scores of 
gifted boys noted what seemed to be an error, a score that was 
more feminine than any for the girls. The score was accordingly 
checked for error, as was also the sex classification of the subject, 
but no error was found. Reference to the field assistant's case 
history revealed the fact that the boy in question (age nine) had 
become a problem to his mother because of his persistent and 
overpowering desire to play the role of a girl. Besides showing a 
distinct preference for feminine plays and games, X frequently 
dressed himself in girl's clothes and "dolled" his face with rouge, 
lipstick, and other cosmetics. When he found that his feminine 
behavior was beginning to attract the attention and disapproval 
of his parents and playmates he cleverly found a way to continue 
it without criticism by writing little plays for neighborhood 
performance, each carefully provided with a feminine role for 
himself. A follow-up study six years later showed that the 
feminine inclinations of X had become more rather than less 
marked. For now, at the age of fifteen, one of his favorite 
amusements was to dress himself as a stylish young woman, apply 
cosmetics liberally, and walk down the street to see how many 
men he could lure into flirtation. It is practically certain that 
at this time X had no knowledge whatever about the existence of 
such a thing as homosexuality. 

So striking a case of inverted behavior in childhood naturally 
provoked speculation in regard to the course of development that 
would ensue. To secure the facts, however, was not easy. 
Letters of inquiry were addressed to the mother from time to 
time, but as these were couched in general terms to avoid the risk 
of causing offense or shock their import was not understood and 



ORIGIN OF THE M-F TEST 15 

they elicited little information. Finally, one of these letters 
which was somewhat more pointed than usual was shown to X 
by his mother. X understood immediately the purpose of the 
inquiry and informed his mother that there was something he had 
kept from her, namely, that his love interests were unlike those 
of other men in that only boys had the slightest attraction for him. 
He denied, however, that he had engaged in any kind of overt 
homosexual behavior. As a result of this confession X was asked 
to fill out the ' ' Attitude-Interest Analysis Test " (M-F test) . The 
score of 71 which he now earned places him near the 50th 
percentile for women, or more feminine than 999 men out of 1,000. 

A few weeks before this chapter was written the mother of X 
requested information about the possibility of normalizing her 
son's emotional life by the use of testoserone, a recently 
developed synthetic preparation of male hormone. 1 

The case we have just described raises in dramatic form the 
question as to the age when an individual's M-F status becomes 
relatively fixed. To what extent is the adult M-F status of a 
subject foreshadowed in the years of middle childhood and 
preadolescence? Does the very feminine boy usually become a 
very feminine man, or is X an exception to the rule? Similarly, 
does the tomboy usually develop into the masculine type of 
woman, or is tomboyishness more commonly but a passing phase? 

The question is a very interesting one, but at present no answer 
can be given. It cannot be answered until a more satisfactory 
M-F test has been devised for use with young children. The 
Plays, Games, and Amusements test is both too crude and too 

1 This youth, now in his twenties, is well started on what promises to be a 
brilliant career in one of the arts. There were several factors which may have 
contributed to the development of his homosexual tendencies. The mother 
married in the late thirties a man some twenty years her senior. When X was 
born he was definitely marked to remain an only child, and the stage was thus 
set for an excessive attachment to his overcherishing mother. Masculine con- 
tacts were largely lacking, as he did not spend much time with other boys and 
had little in common with his elderly father. His artistic temperament and 
refined sensibilities may also have played a part. X is an example of the highest 
type of mental sexual inversion; he has high principles, is passionately devoted 
to his work, and seems to have rejected all overt expression of his homosexual 
inclinations. 



16 SEX AND PERSONALITY 

limited in scope to afford a measure comparable in reliability 
and validity to the M-F test we have devised for adults. Even 
so, the reader will want to know to what extent this test, given 
in the preadolescent years, agrees with scores made by the same 
subjects on the M-F test several years later. 

Six years after the P. G. and A. test was administered to the 
gifted subjects, the M-F test described in this volume was given 
to 94 boys and 99 girls of the same group. The ages at the time 
of the earlier test averaged about 103^ years and ranged from 
6 to 13, giving a mean age of about 16J^ at the time of the 
M-F test. 

In the case of the boys the correlation between the two sets 
of scores was .30 .08, which is large enough to be statistically 
significant but too small to serve as a basis of prediction in the 
case of individual subjects. For the girls the correlation was .24. 
In the case of both boys and girls, however, there were striking 
differences in the M-F scores of subjects who scored at opposite 
extremes on the P. G. and A. test. For example, the 11 most 
masculine boys on the P. G. and A. test had a mean M-F score 
of +78 six years later, as compared with a mean of +37 for the 
6 most feminine. The two M-F scores, +78 and +37, are 
separated by approximately one S.D. of the distribution of a 
typical male group. (See norms, Appendix I.) The 12 most 
masculine girls on the P. G. and A. test had a mean M-F score 
of 50 six years later, as compared with 106 for the 12 who 
had been rated most feminine by the P. G. and A. test. This 
difference is about 1.5 S.D. of a typical female distribution. 

On the whole, these results suggest that there is a certain 
amount of "constancy" with regard to a subject's M-F status 
from middle childhood to the adult period. We are inclined to 
think that improved tests for the earlier years will reveal a 
considerable degree of constancy. The case history data 
presented in Chapters XIII to XV lend considerable support 
to this view. Because of the far-reaching effects which extreme 
M-F deviation may have upon an individual's social and sexual 
adjustment, the problem should be thoroughly investigated. 
In this connection one would like to know whether the child's 



ORIGIN OF THE M-F TEST 17 

progress toward the adult M-F status is related to early or late 
puberty, or influenced by forced association with older or younger 
children as in the case of pupils who are markedly accelerated or 
retarded in school. 

The facts that have been presented, representing, as they 
do, the only examples of subjects being given an objective 
M-F test and later followed on to the adult period, are surely 
challenging enough to justify all the labor that has been expended 
in improving the test, in securing norms for a large number of 
population groups, and in correlating the scores with other 
variables. 



CHAPTER III 
EXTENSION OF THE M-F TEST TO NEW TECHNIQUES 

Each exercise of the M-F test in its present form required a 
large amount of preliminary investigation, including the examina- 
tion of experimental literature in the search for dues, selection of 
tentative lists of items, trial for rejection of non-discriminating 
items, application of retained items to numerous groups of sub- 
jects, computations of reliability and of overlap of sex groups, 
trial of different response methods, and comparison of the merits 
of various scoring and weighting techniques. The reader would 
hardly be interested in a minute account of this exploratory work 
and we shall therefore summarize in a single chapter the 
thousand or more pages of material representing this stage of the 
investigation. 

It was not until the summer of 1925 that opportunity arose to 
provide a broader basis for the M-F test than sex differences 
in the field of plays, games, and amusements. Examination 
of the literature suggested many possible lines of extension, but 
consideration was given only to methods which would permit 
group testing, pencil-and-paper responses, and objective scoring. 
Within a few months experiments were begun with five tech- 
niques: word association, ink-blot association, information, 
interests (likes and dislikes), and tests of introvertive tendencies. 
Two other methods were added to the experiment in 1926, a test 
of emotional and ethical response and a test of common beliefs. 

Apart from a few minor experiments which were quickly found 
to be unpromising, the investigations at this stage were confined 
to the above seven techniques. All of these turned out favorably 
with the exception of the test of opinions, which netted only a 
small proportion of items showing reliable sex differences. 
Although the investigation proceeded upon several lines simul- 
taneously, often with application of the various tests to identical 

18 



EXTENSION OF THE M-F TEST 19 

groups, the account which follows will deal with the seven tech- 
niques successively. 

.THE WORD ASSOCIATION TEST 

The so-called Jree association. method employing words as 
stimuli has been used by numberless investigators since the work 
by Aschaffenburg was first published in 1889. Among these, 
Jastrow, Calkins, Washburn, and others had made note of sex 
differences revealed by the method. 1 The sex differences that 
had been found were usually the by-products of studies primarily 
designed either to establish the general laws of associative 
response or to bring out the influence of individual differences 
other than sex, including personality type, psychotic tendency, 
mental complexes, consciousness of guilt, etc. It was reasonable 
to suppose that if a list of stimulus words was chosen specifically 
on the basis of significant response differences yielded by male and 
female subjects, it would prove a valuable addition to the test of 
masculinity and femininity. 

Dr. J. B. Wyman, in connection with the Stanford study of 
gifted children, 2 had already demonstrated that a word associa- 
tion test can be constructed which is effective in discriminating 
between groups of subjects differing in interests, attitudes, and 
thought trends. Her investigation was not primarily concerned 
with sex comparisons, but with groups representing the extremes 
of intellectual interests, social interests, and activity interests. 
The same principle, however, was involved. If the free associa- 
tion technique is capable of differentiating between subjects 
having much or little of one of three types of interest, it should be 
capable of discriminating between subjects differing in mental 
masculinity and femininity. Dr. Wyman's adaptation of the 
method to group testing, making feasible its application to large 
populations, was also an important consideration. Her pro- 

1 For a summary of the literature on this point see Catharine Cox Miles and 
Lewis M. Terman, Sex differences in the association of ideas, Am. J. Psychol., 
1929, 41, 165-206. 

TERMAN, LEWIS M., et a/., Genetic studies of genius, vol. I, pp. 455-484, 
Stanford Univ. Press, 1925. 



20 



SEX AND PERSONALITY 



cedure was to expose the stimulus word visually, printed in giant 
type on a card 3 by 12 in., and have the subjects write in a 
numbered space~tEe one "word it marfc thpm tgTn^7rr~ The 
responses were scored by the use of weights assigned to them on 
the basis of the frequency difference between subjects rated high 

FORMA 



TABLE 


NAKED 


EVIL 


BLUE 


CASTLE 


ARTIST 


RAIN 


CHEAT 


RADIO 


TURNIP 


ATHLETE 


SAFETY 


BOOK 


FEELING 


BATTLE 


PURE 


FELLOW 


MAIDEN 


KITCHEN 


WAR 


EMBRACE 


FLESH 


BABY 


WASH 


ADMIRE 


PAIN 


ANKLE 


SICK 


MONEY 


WISH 


HAIR 


SQUEEZE 


BULL 


FLIRT 


ACTOR 


WEEP 


MOON 


PROTEST 


WASTE (WAIST) 


MODEST 


KNEE 


HERO 


HUNT 


DEVIL 


PHYSICAL 


FRESH 


ENJOY 


SERVE 


FASHION 


JEALOUS 


ROSE 


DANGER 


SHOP 


MAGIC 


DECENT 


DIMPLE 


SKIRT 


WEDDING 


AFRAID 


MANNERS 


NECKLACE 


LIPS 


PROPOSE 


COLOR 


CLOTHES 


RING 


RELIGION 


ANGER 


BUILD 


HANDSOME 


BREAST 


MOUSE 


VIRTUE 


PROMISE 


LOVE 


WANT 


GENTLE 


FOOT 


SHOULDER 


FAME 


HANDKERCHIEF 


CLOSET 


BOSS 


BOLD 


DUTY 


TOMBOY 


DARLING 


FEMALE 


GIVE 


SCREAM 


PICNIC 


FAITHFUL 




DELICATE 


AUTOMOBILE 




FAIR 


BLUSH 




GARDEN 


FORM 




POWER 


CHILD 




HOME 


BALL 




HAND 


SNAKE 




THIRTEEN 


STOUT 




FILTHY 


POWDER 




SOCIETY 


SWEETHEART 





EXTENSION OF THE M-F TEST 



21 



or low in a given type of interest. A majority of the stimulus 
words in the final form of her test yielded responses which could 
be weighted for two or all three of the types of interest the author 
was attempting to measure. Reliability coefficients found for 
the list of 120 stimulus words ranged from .83 to .87 for intellec- 

FORM B 



CHAIR 


BARE 


DARKNESS 


GREEN 


KNIGHT 


ART 


WIND 


SNEAK 


ELECTRICITY (ELECTRIC. 


CARROT 


STRONG 


SAVE 


LETTER 


MEMORY 


KILL 


TRUE 


MATE 


YOUTH 


CELLAR 


FIGHT 


KISS 


BODY 


BIRTH 


SOAP 


DESPISE 


HUNT 


SLENDER 


ILL 


EARN 


DESIRE 


FACE 


HUG 


ROOSTER 


SPOON 


STAGE 


CRY 


TWILIGHT 


POLICEMAN 


SKIN 


BASHFUL 


LEG 


LEADER 


GUN 


ANGEL 


IDEAL 


CHEEK 


FUN 


SUBMIT 


TASTE 


VAIN 


LILY 


ESCAPE 


BARGAIN 


CHARM 


PROPER 


CURL 


CLOTHING 


MARRIAGE 


FEAR 


RUDE 


DIAMOND 


MOUTH 


ENGAGEMENT 


PINK 


DRESS 


BELL 


WORSHIP 


CURSE 


MAKE 


LOVELY 


HIPS 


WORM 


HONOR 


SUCCESS 


PASSION 


LONGING 


TENDER 


ARM 


THROAT 


AMBITION 


GARTER 


SECRET 


MISTRESS 


DARE 


SACRIFICE 


SISSY 


BELOVED 


MALE 


GIFT 


THRILL 


PARTY 


IMMORTAL 




COARSE 


MACHINE 




BRAVE 


SHAME 




FLOWERS (FLOWER) 


FIGURE 




RULE 


MOTHER 




FAMILY 


PLAY 




LIMB 


TOAD 




SIXTEEN 


THIN 




NASTY 


PAINT 




SOCIAL 


LOVER 





22 SEX AND PERSONALITY 

tual interests, from .82 to .87 for social interests, and from 
.48 to .87 for activity interests. After correction for attenuation 
the correlations of the scores with teacher ratings of the subjects 
for the three types of interests averaged, for six groups of subjects, 
.65 for intellectual interests, .50 for social interests, and .31 for 
activity interests. 

In devising the M-F association test the procedure followed was 
in all essential respects that used by Dr. Wyman. First a short 
EngKslTdiptionary was scanned in the search for words which 
looked 1 ^ though they might bring sex differences in responses if 
tfsecTas stimulus words in an association test. A tentative list 
of about 500 such words was selected by the senior author, with 
the assistance of Edith M. Sprague. The selection was based in 
part on investigational data injthe. field of sex differences, but to a 
greater extent on siitjjectiye "hunches?> Each word in this list 
was next rated by three^u3ges for probable merit in bringing out 
sex differences, with the result that the list was reduced from 
500 to 220 (110 in each of two forms). The 220 stimulus words 
were printed in large type on cards like those used by Dr. Wyman 
and were administered to 400 high-school and university subjects 
equally divided between the sexes. The words were divided into 
two forms, A and B, as shown on pages 20 and 21. 

A few examples of responses found to be predominantly mascu- 
line or feminine which will give an idea of the possibilities of a 
test of this kind in the study of sex differences are shown on 
the following page. 

Reliabilities were secured by application of the test to 128 boys 
and 134 girls in the junior high school and 87 boys and 92 girls in 
the senior high school. Even-numbered items were correlated 
with the odd-numbered and the coefficients were corrected by the 
^Spearman-Brown formula. This was done for scores obtained by 
two methods: (a) by weighting responses from +6 to 6 accord- 
ing to the amount of sex difference shown, and (b) by weighting 
each response as either +1 or 1 (masculine or feminine). The 
*^o sets of scores will hereafter be designated by the terms 
"weighted scores" 1 and "unit scores." For single-sex groups 

1 Sec pp. S2/. for method of deriving score weights. 



EXTENSION OF THE M-F TEST 



23 





Responses given more 




often 


Stimulus word 






By males 


By females 


BLUE.. 


spectrum 


dress 


FLESH. 


meat 


pink 


CLOSET 


door 


clothes 


GARDEN 


weeds 


flower 


HOME. . 


house 


happy 


POWDER 


bullet 


rouge 


CHARM. 


snake 


beauty 


ARM. 


leg 


limb 


FAIR . . . 


weather 


blonde 


RELIGION 


God 


church 


WAR 


soldiers 


hate 


STOUT 


strong 


fat 


GENTLE. 


horse 


mother 


HUNT 


shoot 


find 



the reliabilities (based on 220 words) ranged from .45 to .64 for 
weighted scores, and slightly lower for unit scores. Weighted 
score reliabilities for the sexes combined ranged from .60 to .81. 

An association test of the kind just described has three serious 
disadvantages. (1) A majority of the responses have such low 
frequency that very large numbers of subjects must be tested in 
order to secur^reliable sex differences for a reasonable proportion 
of the responses occurring. This means that of the entire number 
of responses given by a subject, many will be responses which 
cannot be scored. This, of course, seriously reduces the relia- 
bility of the test. (2) Scoring is laborious and time consuming, as 
each of the responses must be looked up in a tabulated list to 
find the weight it carries. (3) Occasionally a subject with defec- 
tive vision fails to perceive correctly jhe. stimute^wordy^while 
others, failing to find a response to a given stimulus word in the 
time allowed (IQ-seeends), lose their place in the column of spaces 
and misplace succeeding responses. Such misplacement is likely 
to cause errors in scoring. 

Because of these defects the type of association test first used 
was finally abandoned in favor of one which required the subject 



2 SEX AND PERSONALITY 

to check tlm^pn^gf^mr jjVm responses , which seemed "to go 
best with" the^ stimulus wordX The stimulus words were no 
longer presented on cards, but were printed in capital letters in a 
test booklet. Each stimulus word was followed by the four 
response words printed in lower case and smaller type. Response 
words were selected which had shown differences in frequency 
between male and female subjects, and of the four following each 
stimulus word two were "masculine" and two "feminine." The 
masculine and feminine responses were arranged in chance order 
to allow for the tendency of some subjects to check more often 
the response standing in a particular serial order. 

An experiment was made to find out whether it is possible to 
make up valid items of this kind without first giving them to 
subjects in the form of a free association test in order to find 
response words that have different frequencies for males and 
females. A list of 51 untried stimulus words was made up, each 
followed by two response words selected as likely to be masculine 
and two as likely to be feminine. Of the 51 items devised in this 
way 28 proved to be usable and in no way inferior to those for 
which response words were selected on the basis of empirically 
ascertained sex differences in frequency. This method saves 
much time and could probably be used advantageously in adapt- 
ing the multiple-response association test to purposes other than 
the study of sex differences. 

The multiple-response form of the association test is better 
adapted to group testing, requires less time for its administration, 
and can be scored with far greater rapidity. However, the reader 
will naturally raise the question whether the test thus altered 
does not lose its essential character as an association technique. 
Certainlj^thejassociation is less " free^'-since response is limited to 
choke among the four alternatives- presented. Even if these 
have been selected from among responses showing high frequency 
by the earlier method it will nevertheless often happen that in the 
case of a particular subject no one of the four would have been 
given by the method of free association. However, the relative 
merits of the two methods can only be judged by their results. 
Our data show that by every criterion the multiple-response 



EXTENSION OF THE M-F TEST 25 

method is as good for the present purpose as the older method. 
It is fully as reliable, gives as wide separation of sex groups, and 
in fact correlates with scores obtained by the standard method 
almost as highly as the latter correlate with themselves. 

The experimental form of the multiple-response association test 
was composed of 171 items, 120 from the 220 of the original list 
and 51 items artificially constructed. It was given to 600 
subjects: 100 of each sex in the seventh grade, the junior year of 
high school, and the university. Both weighted and unit scores, 
based upon the sex differences found by item tabulation, were 
used in computing reliabilities and sex overlap. The weighted 
scores showed only a slight superiority when judged by these two 
criteria and were discarded in favor of the simpler unit scores. 1 

All of the types of association tests we have used as measures of 
M-t differences have low reliability. It will be recalled that the 
free association list of 220 stimulus words used in the first experi- 
ment had a reliability of only .45 to .64 for single-sex groups, and 
of only .60 to .81 for the sexes combined. The reliabilities of the 
multiple-response association test, with 171 stimulus words, 
averaged .59 for single-sex groups and .78 for the sexes combined. 
TEe amount of sex overlap was close to 10 per centloFeach of the 
three groups tested. 2 

For the final M-F test 120 words were selected from the 171 
used in the experiment just described. These have been equally 
divided between Form A and Form B and are reproduced in 
Appendix IV. The words retained include only those from the 
original list which showed sex differences in the same direction 
for at least three out of the four multiple responses in all the 
groups tested. The reliabilities were then computed for the 
retained lists of 60 words, both by the split-half method and by 
the correlation of Form A against Form B. Average reliabilities 
did not differ significantly for the two methods, averaging about 
.40 for single-sex groups and .55 for the sexes combined. By 
application of the Spearman-Brown formula the reliability of the 

1 See pp. 53/. for discussion of considerations leading to this choice. 
'Throughout the M-F study the percentage of overlap has been computed 
by the method described on pp. 66 Jf. 



26 SEX AND PERSONALITY 

120 words of Form A and Form B combined is estimated to be 
about .57 for single-sex groups and .71 for the sexes combined. 
These figures would have been somewhat higher if less homogene- 
ous groups had been used, that is, if subjects from junior high 
school, senior high school, college, and a wide variety of adult 
populations had been thrown together. Even go,- the reliability 
of a test of this type tends to be low. The score on Exercise 1 
should not be used by Itself for the comparison of individuals 
unless both forms (120 words) have been given, and then only 
with due caution. It is estimated that a test of nearly 300 words 
would be required to give aTeliabihty oi .75 ToFa single-sexjyroup 
of .85 for the sexes combined. 

The amount of sex overlap based upon 60 words was 22 per cent 
in the seventh grade, 14 per cent in the high school, and 12 per cent 
for the college subjects. Overlap has not been computed on the 
basis of 120 words, but it would of course be appreciably less. 

Score norms for Exercise 1 are given in Appendix II. We 
are here concerned only with the outstanding facts brought out 
in the adaptation of the word-association technique in the study 
of M-F differences. We feel warranted in concluding that the 
test has demonstrated its value for this purpose when used as one 
of a battery of tests, though in the interest of reliability the 
number of items should be considerably increased if it is to be 
used independently in profile studies of individual subjects. 

INK-BLOT ASSOCIATION 

This form of the association test has an extensive history 
extending from the time it was suggested by Binet in 1895 to 
recent and current work with the Rorschach test. Dearborn, 
Sharp, and Bartlett have used the method in qualitative studies 
of the imagination; Kirkpatrick and Pyle in investigations of the 
associations of school children; Brown as a test of suggestibility; 
Rorschach and his followers as a test of personality type. 1 
Extravagant claims have been made with respect to its value in 

1 See WHIPPLE, G. M., Manual of mental and physical tests; Part II, Complex 
processes, pp. 254 Jf., Warwick and York, 1915. 



EXTENSION OF THE M-F TEST 27 

the classification of abnormal types, but as a test of a somewhat 
"freer" kind of association than is called forth by word stimuli it 
would seem to have a distinct place in the investigation of per- 
sonality differences. A disadvantage is that so many associations 
are possible with the usual kind of ink blot that it is very difficult 
to devise a scoring technique which gives high enough reliability 
to make the test worth using. 

The ink-blot test in the study of M-F differences began in 1925 
when the senior author and Mary A. Bell produced a series of 
40 blots according to the directions of Dearborn. 1 In this case 
blots were so made as to suggest responses which it was thought 
might show sex differences. For example, one blot was made to 
look something like a hill and also something like a colonial hat, 
on the theory that the latter association would be more often 
given by women than by men if, as is commonly believed, women 
are more interested than men in wearing apparel and less in the 
kind of outdoor life that would suggest the response "hill." 
Another was devised to resemble both a pistol and a cradle, with 
the idea of tapping man's pugnacious and woman's maternal 
attitudes; and so with the entire series. The trial series also 
included the 20 blots used by Whipple. 

Tryout of these 60 ink blots with men and women college 
students brought to light various imperfections, the most serious 
of which was the wide range of responses with consequent low 
frequency for the greater number and low reliability of sex 
differences. It was found that stimuli with simple outlines were 
better suited to the purpose at hand than those that were more 
complex and irregular. The first series was therefore discarded 
and a new series of 100 drawn in printer's heavy ink with a paint 
brush wielded in rather sweeping strokes. By this method it was 
easier to devise stimuli offering a more strictly dichotomous 
ambiguity. For example, one blot might very well be taken 
either as a pipe or as a mailbox, one as a snake or a tulip, one as a 
snowshoe or a spoon, one as the moon or a ship, etc. 

1 DEARBORN, G., A study of imagination, Am. J. Psychol., 1898, 9, 183-190. 
Also WELLS, F. L., Rorschach and the free association test, /. Gen. Psychol., 
1935, 13, 413-433. 



28 SEX AND PERSONALITY 

After trial of the new series with 100 male and 100 female stu- 
dents in high school and college, 30 of the 100 were discarded. 
Photographic plates were made of those retained and the blots 
were printed in a 3 by 5 in. booklet of 70 leaves, one blot to a leaf. 
On the outside cover were the following directions: "On each leaf 
of this booklet is a kind of ink blot or drawing. They are not 
pictures of anything in particular but might suggest almost 
anything to you, just as shapes in the clouds sometimes do. 
Below each drawing write the first thing it makes you think of." 
At intervals of ten seconds the subjects were instructed to turn 
the leaf to the next drawing. 

The test in this form was given to 460 subjects in the following 
groups, each evenly divided between males and females: 120 
seventh-grade pupils, 180 high-school freshmen, 80 students in 
the State Teachers College of San Jose, and 80 nonacademic 
adults. The last group included 40 members of a professional 
and business women's club and 40 members of a men's church 
club. The responses given by these subjects were tabulated 
and the frequencies were computed separately by sex. Only 
those responses were retained for scoring which were given by as 
many as four subjects in at least three of the four groups tested 
and which showed a sex difference in the same direction for at 
least three of the four groups. Each retained response was then 
assigned a weight in proportion to the standard error of the 
difference between the percentage of males and females giving it. 
The weights ranged from to IS. The total score of a subject 
taking the test was computed by taking the average weight of his 
scorable responses. The average of the weights, instead of their 
sum, was used because the number of scorable responses varied 
from subject to subject. The score distributions of 230 male and 
230 female subjects, together with their accumulative per cents, 
are given in Table 3. 

The reliability of the 70-blot test, computed by the split-half 
method, for the 180 high-school freshmen was found to be .64 
for the sexes combined but only .48 and .31, respectively, for 
males and females taken separately. The validity of the test is 
indicated by the smallness of sex overlap shown in Table 3. 



EXTENSION OF THE M-F TEST 29 

TABLE 3. DISTRIBUTION or INK-BLOT SCORES or MALES AND FEMALES 



Score 


Male 


Female 


Cumulative 
per cent 
males 


Cumulative 
per cent 
females 


7. 


1 




100 




6 2 


1 




99 6 




6 1 






99 1 




6 


2 




99 1 




5 9 


2 




98 3 




5 8 


6 




97 4 




5 7 


4 




94 8 




5 6 


10 




93 




5 5 


14 


1 


88 7 


100. 


5 4 


7 




82 6 


99 6 


5 3 


20 




79 6 


99 6 


5 2 


25 




70 9 


99.6 


5 1 


14 


1 


60 


99 6 


5 


25 




53 9 


99.1 


4 9 


17 


2 


43 


99 1 


4 8 


15 


4 


35 7 


98.3 


4 7 


14 


7 


29 1 


96.5 


4 6 


15 


14 


23 


93.5 


4 5 


11 


10 


16 5 


87.4 


4 4 


8 


17 


11.7 


83 


4 3 


7 


13 


8 3 


75 7 


4 2 


5 


25 


5 2 


70 


4 1 


4 


22 


3 


59 1 


4 




21 


1.3 


49 6 


3.9 




24 


1 3 


40 4 


3.8 


2 


29 


1.3 


30.0 


3 7 




17 


.4 


17 4 


3 6 




6 


.4 


10 


3 5 




7 


.4 


7 4 


3.4 


1 


3 


.4 


4 3 


3.3 




2 




3 


3 2 




1 




2.2 


3 1 




2 




1.7 


3 








8 


2 9 




2 




.8 


Total 


230 


230 







30 SEX AND PERSONALITY 

In a later edition of the test the number of blots was reduced to 
50 by elimination of those which had been found least satisfactory. 
As in the case of the word-association test, the original procedure 
was discarded in favor of the multiple-response technique. All 
the considerations which favored this change in the test of word 
association apply with equal force to the ink-blot test. For each 
item four responses were selected from those secured by the free 
association method, two which had been given more often by 
males and two more often by females in at least three of the four 
groups of subjects, all with as high frequencies as could be found 
satisfying this criterion. 

The SO blots were printed in a booklet with six to eight on a 
page. Beside each blot were the four responses arranged in 
chance order in a column. The first page contained the instruc- 
tions and two blots used as a fore-exercise. 

The test was given in this form to 600 subjects, 200 in the eighth 
grade, 200 in the junior year of high school, and 200 in college. 
Each group contained 100 males and 100 females. Reliabilities 
and sex overlap were computed separately for weighted and unit 
scores. The superiority of weighted scores was so slight by both 
criteria that they were abandoned in favor of the more readily 
computed unit scores. By both methods of scoring the reliability 
for SO items is low and the overlap of sexes large. An average 
reliability of .43 was found for the sexes combined; the reliabilities 
for single-sex groups ranged from .00 to .35. 

In the final edition of the M-F test the number of blots was 
further reduced to 36, equally divided between Form A and 
Form B. Reliabilities found for eight new groups, each number- 
ing 50 subjects of one sex, ranged from almost zero to .56 for the 
36 items. The 18 items in a single form have an average relia- 
bility of .28 for groups homogeneous as to sex and school grade, 
with consequent heavy overlap between the score distributions 
for male and female subjects. The test could therefore have 
been eliminated from the battery without appreciable loss, but 
has been retained partly because of its appeal to subjects and 
partly in the hope that its retention would stimulate further 
experimentation. It seems in fact to have all the merits possessed 



EXTENSION OF THE M-F TEST 31 

by the word-association method and possibly others. Certainly 
it would seem to evoke associations less stereotyped by habit than 
those evoked by words. In order to make the test produce a 
reasonably reliable score it would of course be necessary to 
increase the number of stimuli to something like 200. As this 
test stands its score should not be used by itself for the comparison 
either of individuals or of moderate-sized groups. 

THE INFORMATION TEST 

Tests of general information of the type commonly used in 
group tests of intelligence usually show relatively small sex 
differences in total score. This seems to be explained by the con- 
scious intent, on the part of those who devise such tests, to make 
them equally "fair" to male and female subjects, or, when this is 
not deemed feasible, to select items equally divided between 
those which favor males and those which favor females. The 
Army Alpha information test is an exception; its items were 
selected for testing only male subjects, with the result that when 
otherwise comparable sex groups are given this test the mean 
information score is usually found to be lower for females. There 
is abundant evidence in the literature of sex differences that 
particular items of information are more often answered correctly 
by males and others by females, and it was this fact which gave 
rise to Exercise 3 of the M-F test. 

Early in 1926 the literature on sex differences in information 
was canvassed by the senior author and the first M-F information 
test of 200 items was made up. The selections were based in part 
upon the direct evidence offered by tabulations of successes and 
failures by males and females on various information tests, and in 
part on indirect evidence from sex differences in interests, reading 
preferences, etc. The 200 items may be roughly classified as 
follows: 28 historical; 26 from the physical sciences; 23 from the 
biological sciences; 35 literary (relating to books, authors, and 
literary characters); 17 items of general information, mostly of a 
practical kind; 38 related to household arts; 16 from religion and 
mythology; and 17 miscellaneous. 



32 SEX AND PERSONALITY 

The items were set up in the multiple-response form, the sub- 
ject's task being that of underlining the correct response out of 
four given. The test was administered to 800 subjects, 100 of 
each sex in each of four populations: seventh-grade pupils, high- 
school students, college students, and nonacademic adults. The 
successes and failures on each item were tabulated separately by 
sex in each of the four populations, and only those items were 
retained which showed a probably significant sex difference in 
the same direction for at least three of the four populations. Of 
the 200 items, 91 satisfied this criterion. These were assigned 
M-F weights in proportion to the amount of sex difference. 

The reliabilities computed for the eight single-sex groups 
(N = 100 each) of the original populations tested averaged .50 
for single-sex groups and .71 for the sexes combined. Unit scores 
and weighted scores gave about the same reliability. 

The 91 items were then administered to new groups, as follows: 
127 eighth-grade boys, 134 eighth-grade girls, 88 high-school 
boys, 122 high-school girls, 212 college men, 170 college women. 
Reliabilities for these groups, based on weights derived from the 
four populations previously tested, averaged .41 for single-sex 
groups and .69 for the sexes combined. These reliabilities are 
much lower than those found when the same items are used simply 
as a measure of information. The latter was found to be .89 
for high-school girls and .79 for high-school boys, as compared 
with about .40 when scored for masculinity or femininity of 
response. 

In an effort to improve the reliability, a search was begun for 
new information items which would yield sex differences. Data 
were tabulated from an information test of 491 items which had 
been administered to a total of 200 subjects of each sex in grades 
8, 11, and 12. The experiment yielded 95 additional items which 
seemed to merit further trial. This new set was given to 121 boys 
and 112 girls in the eighth grade, and to 102 boys and 118 girls 
in the high school. In the further interest of reliability a new 
method of scoring was tried out. Previously a subject's total 
score was the number of masculine items correctly answered 
minus the number of feminine items correctly answered. It was 



EXTENSION OF THE M-F TEST 33 

found, however, that in the case of nearly all the items there was 
a significant sex difference in the number of subjects underlining 
one or more of the three wrong responses, and also in the number 
failing to answer at all. It was accordingly found advantageous 
to score wrong responses and omissions, instead of mere success 
and failure. By this "all-possible-response" method of scoring 
the average reliability for eight single-sex groups of SO subjects 
each was raised to .70, and for sexes combined to .82. These 
figures are for the populations from which the unit weights were 
derived and so are somewhat spuriously high, but the superiority 
of the "all-possible-response" method is unmistakable. 

The final M-F information test was made up of the best 140 
items selected from the two series above described containing, 
respectively, 91 and 95 items. These 140 are equally divided 
between Form A and Form B of the test, the items of each series 
being arranged in order of frequency of correct response. Unit 
scores are used, based upon the " all-possible-response " method. 
The reliability of 70 items varies according to the heterogeneity 
of the groups tested, but is in the neighborhood of .45 for a single 
sex and .65 for the sexes combined. It is only by using both 
forms, totaling 140 items, that the reliability of Exercise 3 taken 
by itself reaches a figure which justifies more than a very tenta- 
tive rating of the individual subject. 

The validity of the 70-item test is indicated by sex overlap 
ranging from 12 to 23 per cent in the various groups tested, the 
average being about 16 per cent. 

Our final appraisal of the information test as an approach to the 
study of M-F differences is that it has demonstrated its value 
when used in conjunction with other tests. Its rather low 
reliability can be raised to any figure within reasonable limits by 
increasing the number of items. However, like all the other 
tests in the battery, it is capable of measuring only one aspect of 
sex difference, whereas measures of several aspects are necessary 
in order to secure a more balanced picture of an individual's M-F 
status. The amount of one's information depends upon school- 
ing and intelligence; its content depends largely upon interests. 



34 SEX AND PERSONALITY 

That sex differences in information are perhaps almost wholly 
due to environmental causes (granting sex equality in general 
intelligence) does not seem to us a valid criticism of the test as we 
have used it. It remains true that, generally speaking, male and 
female subjects acquire funds of information which differ con- 
siderably in content, and it must be held significant when the 
information of a given man or woman includes a disproportion- 
ately large amount of those kinds that are more frequently 
encountered among members of the opposite sex, whatever 
causes may be responsible for the departure from norm. It will 
be granted that these are probably in the main environmental; 
certainly there are many kinds of information exposure to which 
is unequal for the two sexes. We have sought to minimize the 
influence of this constant environmental factor in two ways. In 
the first place we purposely excluded from our trial series items 
to which both sexes have not had reasonable exposure in school, 
home, or other environment. It was hoped that by this means 
whatever differences were found would be more indicative of 
interests or temperament than would otherwise be the case. 
Secondly, the method of unit scores avoids giving undue weight to 
the very large sex differences which might reasonably be held to 
reflect largely differences in environment. By the unit method 
an item showing the largest sex difference contributes no more 
toward the total score than one showing a barely significant 
difference. The amount of sex overlap by this method is 
increased, but we believe that the psychological significance 
of the score has been enhanced. 

A special advantage of including a test like Exercise 3 in the 
M-F battery lies in the fact that it offers a rough measure of 
intelligence when it is also scored as a straight information 
test. The intelligence score thus derived has a reliability (Form 
A against Form B) of .77 and yields a correlation of .67 with 
mental age on the Terman Group Test for eighth-grade subjects. 
It predicts M.A. on the T.G.T. with a probable error of 12 months. 
Although only a very rough measure of the intelligence level of a 
single individual, it is sufficiently reliable and valid to be very 
useful in the comparison of groups. 



EXTENSION OF THE M-F TEST 35 



TEST or EMOTIONAL AND ETHICAL ATTITUDES 



s no other differences between the sexes have been so 
much stressed as those in the field of emotions. Psychologists, 
psychiatrists, sociologists, novelistSj^ramatists, and poets are in 
fairly general agreement that thf affective ufe of women tends in 
a number of ways to be unlike that of men. In the first place, 
women are characterized as having a higher degree of affectibility 
than men in the sense that they respond emotionally to a greater 
variety of stimuli and experience more intense feeling. Secondly, 
it is commonly believed that there are important sex differences 
in the relative contribution of the specific emotions to the total 
affective life; man, for example, is supposed to be more given to 
anger and woman to the sympathetic emotions. Thirdly, sex 
differences are alleged with respect to the stimuli that are effective 
in calling forth a given emotion; men and women do not fear the 
same things, are not angered by the same things, do not pity 
the same people or withdraw in disgust from the same objects. 
Finally, there are thought to be qualitative as well as quantitative 
differences in the emotional experiences of the sexes. It may 
well be that a man's fear or anger or love does in fact have a 
different existential quality from the corresponding emotion of a 
woman, but whether this is true is a question that can never be 
answered in terms of objective data; introspective testimony is 
at best only able to suggest clues as to the nature of qualitative 
differences that may exist. 

The first three types of sex difference mentioned above are 
amenable to scientific investigation, and there exists already a 
respectable body of experimental literature dealing with them. 
A survey of this literature suggested that a contribution could 
best be made to the study of M-F differences by asking subjects 
to rate the effectiveness of specific stimuli in provoking them to a 
particular emotional response. Anger, fear, disgust, and pity 
were the emotions chosen for investigation... . " 

The experimental edition of the test was composed of 218 
items: 35 relating to anger, 45 to fear, 41 to disgust, 33 to pity, 
and 64 to ethical attitudes. The items on ethical attitudes were 



36 SEX AND PERSONALITY 

included on the theory that responses of the kind in question are 
fundamentally emotional rather than intellectual. In form the 
test was in all essential respects identical with Exercise 4 of the 
present M-F battery. The subject checked VM, M, L, or N 
(very much, much, little, none) to indicate the extent to which 
a given situation tended to provoke in him the emotion in 
question. The ethical attitude items were responded to by 
checking 3, 2, 1, or to indicate the degree of moral seriousness 
of a given type of behavior. 

Herejjis Jn the case of the other trial series of M-F tests, the 
selection of items had to be based chiefly on subjective opinion 
as to ways in which the sexes are likely to differ. First, several 
persons working inHpppnffcntly, jnrlnding both professors and 
graduate students of psychology, made up lists of stimuli which 
in their opinion tend to rouse more anger (or fear, etc.) in one sex 
than in the other. Next, conferences were held for discussion of 
individual items proposed, in order to eliminate those not favored 
by at least a strong minority of opinion. In that part of the 
test dealing with ethical attitudes it was possible to base the 
selection of items largely on the data of earlier investigations by 
others. 

The trial series of 218 items thus selected was administered to 
854 subjects: 148 boys and 172 girls enrolled in the eighth grade, 
116 boys and 128 girls in high school, and 144 men and 146 women 
attending college. Responses were tabulated separately by sex 
for each of the three populations tested and scoring keys were 
made for computing both weighted and unit scores. Items which 
did not yield a significant sex difference on at least two of the four 
multiple responses (VM, M, L, N or 3, 2, 1, 0) were disregarded 
in scoring. In this experiment a total score was computed for 
each of the four parts of the test; that is, for anger, fear, disgust, 
pity, and ethical attitudes. 

The emotional-attitude test proved successful beyond all 
expectations. In the first place, approximat^y 90 pffr cent of the 
items in the trial series were sufficiently discriminative to justify 
their retention, whereas jjx^tSe case of several of the other tests 
discards from tEe trial series ran as high as 60 to 75 per cent. In 



EXTENSION OF THE M-F TEST 37 

the second place, the reliabilities were surprisingly high even when 
computed separately for the four parts of the test. These 
(based on unit weights) averaged as follows for six single-sex 
groups: anger, 34 items, .74; fear, 40 items, .81 ; disgust, 36 items, 
.83; pity, 29 items, .86; ethical attitudes, 56 items, .88. The 
reliabilities based on weighted scores were of about the same 
magnitude as those found for unit scores. Sex overlap was 
approximately 30 per cent for anger, fear, and disgust, and 
35 per cent for pity and ethical attitudes. 

The 195 items retained were divided between Form A and 
Form B of the final M-F test. In making the division care was 
taken to balance the A and B lists as accurately as possible both 
for content and for comparability of means and score ranges. 
As the number of items available for one form of the test (A or B) 
is only 15 to 28 for each section of the test, the four sections are 
combined into a single total score. Both forms of the test as 
thus made up were given to new groups of subjects in the eighth 
grade and high school and reliabilities of total score on Form A or 
Form B were computed for eight single-sex groups of 50 subjects 
each. The average was .88. Reliability of total score for mixed- 
sex groups is .90 or better by the same method (split-half). 
These reliabilities are remarkably high for a test composed of less 
than a hundred items. Sex overlap on total score ranges from 
21 per cent to 34 per cent, with little difference between Form A 
and Form B. 

We have been somewhat puzzled to account for the high relia- 
bility of this test as compared with the association and informa- 
tion tests. It appears that 20 items in any division of the test 
yield as consistent results as several times that number of associa- 
tion or information items, and that the total score based upon 
less than 100 items is as reliable as the best tests of intelligence 
requiring an hour for their administration. One explanation 
may be the greater homogeneity of the test items. The items 
of the association and information tests constitute an extremely 
heterogeneous and small sampling of a vast multitude of possible 
sex differences. The number of situations to which we are likely 
to respond with anger, fear, etc., is much more limited than our 



38 SEX AND PERSONALITY 

associations with a given list of word stimuli, or than the countless 
bits of factual information which one may be more or less likely 
to possess than the average person of his sex. That heterogeneity 
of content does ordinarily tend to lower reUabiKty^a2est4S a 
weU known faet. One sees its effect in tie total score of the 
emotional-attitude test, for the reliability of this total is lower 
than would be predicted by the Spearman-Brown prophecy 
formula from the reliability of any one of the four sections taken 
singly. 

Another factor contributing to reliability is probably the 
graded type of response. A subject's response indicates not 
whether a given situation arouses in him a given emotion, but to 
what extent it does so. Moreover, any tendency a subject may 
have to high or low emotivity would operate throughout the test 
as a constant factor making for increased consistency. The same 
effect would be produced, regardless of emotivity differences, by 
a ^ubject's strictness^or looseness of standard in the use of the 
response words. One who is given to careless use pf superlatives 
may be expected to check VM (very much) with unusual fre- 
quency in all parts of the test. The reliability of the test in 
question is perhaps fully accounted for by homogeneity of content, 
graded responses, and a constant factor of high or low emotivity 
(or an equivalent constant factor in language usage). 

At this point attention may be called to a persistent sex differ- 
ence in the frequency with which male and female subjects check 
VM or N in the first four parts of the test and 3 or in the section 
dealing with ethical attitude. For example, VM and 3 together 
receive a masculine score in the case of only 9 of the 195 items 
in thetwo forms of the test, as compared with 152 which receive a 
feminine weight and 34 no weight at allj On the other hand, N 
and together havejynasculine weight in 133 items as compared 
with 12 which are feminine and 40 which receive no weight 
The question whether women are in fact so much more emo- 
tional than men cannot be answered by the data at hand^their 
appearance of being so may be entirely an artifact of a sex 
di^erencie Tn language usage. That the latter explanation can 
alone be responsible for the results hardly seems to us probable, 



EXTENSION OF THE M-F TEST 39 

especially as the ethical-attitude items show about the same sex 
difference as those dealing more strictly with the emotions. 
Since there is abundant outside evidence that in our culture 
women do in fact tend to be more severe than men in their judg- 
ment of offenses, the sex difference in this section of the test 
cannot be entirely one of language usage. One might infer that 
if language is not the sole factor here it probably is not in the other 
four sections of the test. 

TEST OF INTERESTS 

We have already seen how the conception of a masculinity- 
femininity test had its origin in the treatment of data relating to 
interest in and practice of plays, games, and amusements in a 
study of gifted and control groups of school children. When an 
extension of the M-F test was undertaken one of the first things 
done was to make a collection of items sampling a wider variety 
of interests than the original test. As a result of rather thorough- 
going search of the literature of sex differences in interests, 456 
items were assembled. About three-fourths of these were new 
items; the remainder were selected from the 1924 edition of the 
Strong-Cowdery 1 test of occupational interests after giving the 
test to SO adults of each sex and tabulating the responses for the 
individual items. 

The 456 items were made up in two equivalent forms of 
228 items each. In the case of each item, the subject's task was 
to check L, D, or N to indicate "like," "dislike," or "neither." 
The items were listed in the following groups, each form con- 
taining half the items of a given group: occupations, 78; people, 
72; games and amusements, 78; movies, magazines, and school 
studies, 56; books and literary characters, 68; travel and sight- 
seeing preferences, 52; special interests, 52. This edition 
of the test was given to the 245 subjects listed on the following 
page. 

Item responses were tabulated by sex and only those were 
retained for scoring which showed sex differences that were 

1 COWDERY, KARL M., Measurement of professional attitudes, /. Person. Res., 
1926,^131-141. 



40 



SEX AND PERSONALITY 





Form A 


Form B 


Male 


Female 


Male 


Female 


Seventh grade 
High school 
College... 
Non academic. 

Total 


79 
100 
26 
40 


79 
100 
26 
40 


46 
94 
54 
38 


46 
94 

54 
38 


245 


245 


232 


232 



large enough for two of the three responses (L, Z), or N) to be 
probably significant and that were in the same direction in all 
four groups tested. The reliability of one form of the test 
(A or B) averaged .55 for single-sex groups and .84 for sexes 
combined. The sex overlap was slightly less than 10 per cent. 
These figures are for unit scores and agree very closely with those 
based upon weighted scores. 

A selection was made of the best 170 items from the 456 of the 
original set. They included the following: occupations, 30; 
people, 18; movies, magazines, and school studies, 26; games and 
amusements, 27; books and literary characters, 36; special 
interests, 33. These were administered to 127 males and 134 
females in the seventh grade, to 88 males and 122 females in the 
first year of high school, and to 177 males and 156 females in 
college. For single-sex groups the average reliability (unit 
score) was .70, and for the sexes combined .86. The superiority 
of this selected list of items over the trial series is seen in the 
higher reliability and in the decreased overlap of sex groups 
(6 per cent as compared with 10 per cent). 

Shortly after the above experiment was completed a trial 
was made of 60 newly assembled items relating to historical 
characters. Four populations were tested, each containing 
50 males and 50 females. The reliability of these 60 items 
averaged .83 for the eight single-sex groups higher than had 
been found for much longer lists made up of items from a number 
of widely different fields. This is another illustration of the 
effect of homogeneity of content upon a test's reliability. Fifty- 



EXTENSION OF THE M-F TEST 41 

five of the historical items discriminated sufficiently between 
male and female subjects to justify their retention, 28 being 
assigned to Form A and 27 to Form B of the present M-F test. 
Though logically belonging in Exercise 5, which contains the 
other items on likes and dislikes, these items were located in 
Exercise 6 with a group of others relating to opinions. 

Another group of 40 interest items tried out at this time should 
be mentioned here. In these the subject was asked to give his 
preference between two kinds of activities or situations such as: 
(a) camping, (i) living in good hotels; (a) a well cooked meal 
with soiled linen, (6) a poorly cooked meal with linen spotless. 
The IS items of this list which gave significant differences 
between 239 male and 280 female subjects were printed in the 
present M-F test with Exercise 4, although, like the historical 
items, they logically belong in Exercise 5. 

Exercise 5 as it now stands in the test contains 187 items, 93 in 
Form A and 94 in Form B. They include 50 relating to occupa- 
tions, 24 to people, 27 to movies, magazine reading, and studies, 
40 to games and activities, and 26 to books. The reliability of 
one form, based on eight single-sex groups each numbering 
50 subjects, ranged from .54 to .79 with an average of .66. For 
mixed-sex groups the reliability of Form A against Form B is 
.86 to .88. The overlap of sex groups ranges from 6 to 12 
per cent. 

The test of interests and attitudes has proved to be one of the 
most valuable in the entire M-F battery, as there are several 
types of populations which it especially helps to differentiate 
from the generality of a given sex. Among these are male (and 
probably also female) homosexuals. It would have been 
strengthened, however, by the inclusion of the items in the first 
part of Exercise 6 dealing with historical characters, and the 
small group of items at the end of Exercise 4 on preference 
between two activities. For profile studies it would be desirable 
to regroup the items in this way, also to give both forms of the 
test in order to enhance reliability. The scores so obtained 
could be counted upon to yield reliability coefficients around 
.85 for single-sex groups and dose to .95 for the sexes combined. 



42 SEX AND PERSONALITY 

In view of the fact that more than a third of the total number 
of "interest" items in the M-F test are identical with or similar 
to items in Strong's test of vocational interests, it occurred 
to the senior author that it would be desirable to derive a mascu- 
linity-femininity scoring key for the Strong test to supplement 
the keys giving measures of occupational interest. Arrangements 
were made in 1932 for this to be done by Dr. Harold Carter. 
With the assistance of Dr. Strong, 114 blanks of males who had 
taken the vocational interest test were matched with those of 
114 females and item tabulations were made. Of the 420 items, 
156 were found which could be classified as masculine or feminine 
and these were weighted in proportion to the sex differences 
found. The total score so derived had a reliability of .86 for 
males, .78 for females, and .94 for the combined groups of male 
and female subjects used in deriving the weights. The figures 
would of course be lower for new groups. It was found to 
correlate with the present M-F test to the extent of .43 for a 
group of 41 males and .62 for a group of 62 females. These 
items do not constitute a satisfactory substitute for our M-F 
test, but they make possible interesting comparisons between 
groups which have been given the vocational interest test and not 
the M-F test. Data are presented elsewhere (pp. 117 Jf.) on the 
relation of the occupational M-F score to marital compatibility. 

Dr. Strong has carried further the work initiated by the senior 
author and Dr. Carter by utilizing the data on sex differences for 
603 adults of each sex who had taken his vocational interest 
test. The M-F test which he has derived from these data 
includes 202 of the 420 items in his vocational interest test. 
This form of the Strong M-F test will be found very useful in 
investigations dealing with sex differences in the restricted field 
of interests, but like its predecessor it has too little in common 
with our M-F test to serve as an adequate substitute for it, the 
coefficient of alienation between the two being somewhere in 
the neighborhood of .80, if we base the calculation on the inter- 
correlations for female subjects, and about .90, if based on the 
intercorrelations for male subjects. In other words, the two 
tests are not measuring the same thing to the extent of more 



EXTENSION OF THE M-F TEST 43 

than 10 or 20 per cent. Basing the calculations on intercor- 
relations corrected for attenuation would increase the 10 or 
20 per cent to something like 20 or 40 per cent. It is obvious, 
therefore, that even if both tests had a reliability of 1.00 they 
would still be measuring traits that have too little in common to 
warrant the use of either measure as an equivalent of the other. 1 

A TEST or OPINIONS 

Women are frequently depicted in literature and in the 
popular press as holding opinions different from those of men on 
innumerable matters. The belief that such is the case seems 
to be fairly general, for no doubt as to its truth was expressed 
by a single man or woman of many who were interviewed on the 
question. "Men believe so many things which we women know 
are not true" was a formulation encountered a number of times; 
"Women have such silly opinions about a lot of things" was the 
way men sometimes expressed it. The M-F information test 
had shown that the factual repertories of men and women are 
not the same, and one would naturally infer that this would 
inevitably result in sex differences with respect to opinions held. 
Studies of superstitious beliefs had presented considerable 
evidence of sex differences in that particular realm of belief. 
Moreover, if belief is in part an emotional response, as some have 
held, 2 one would expect this to produce sex differences in the 
opinions which are accepted as true or rejected as false. 

The task of devising a test of opinions was undertaken hope- 
fully and 201 items were assembled, of which 25 were taken from 
Garrett and Fisher's study of 140 women and 219 men in Colum- 
bia University, 3 and 25 from Jones's study of the opinions of 
college students. 4 The remaining 150 items were new. From 

1 STRONG, EDWARD K., JR., Interests of men and women, /. Soc. Psychol., 
1936, 9, 49-67. 

1 See, for example, McDouGALL, WILLIAM, Belief as a derived emotion, Psychol. 
Rev., 1921,28, 315-328. 

8 GARRETT, H. E., and T. R. FISHER, The prevalence of certain popular mis- 
conceptions, /. Appl. Psychol., 1926, 10, 411-420. 

4 JONES, VERNON, Disagreement among teachers as to right and wrong, Teach 
Coll. Rec., 21, 34-36. 



44 SEX AND PERSONALITY 

the entire number, 96 were selected by the composite ratings of 
several judges as most likely to prove valid for the purpose at 
hand. 

The test was given to somewhat more than 100 of each sex 
in the seventh grade, to the same number in the high school, and 
to 50 of each sex attending college. The task as set forth in the 
instructions was to read each statement and " consider whether 
it is mostly true or mostly false/' and to encircle TorF accordingly. 

The results of the response tabulations were disappointing, 
for only 28 of the 96 items yielded sufficiently large and con- 
sistent sex differences in the three populations to warrant their 
inclusion in the M-F battery. These have been divided equally 
between the two forms and appear as part of Exercise 6. Relia- 
bility of the 28 items separately computed for nine single-sex 
groups of approximately 50 subjects each averaged only .40. For 
all the subjects below college level, sexes combined, the reliability 
was .68. These figures are based upon unit scores, though 
weighted scores were tried out and discarded as in the case of 
all the other M-F studies. Overlap of sex groups ranged from 
22 to 28 per cent. 

The outcome of the experiment was so meager that it is 
obvious the test of opinions could have been eliminated from the 
battery without serious loss. It was finally retained in the 
interest of making the total score of the M-F battery a wide 
sampling of sex differences. Even when both forms of the opin- 
ions test have been given their combined score is still entirely 
too unreliable to justify its separate use except in the com- 
parison of exceedingly large populations. 

What conclusions can be drawn with respect to the importance 
of beliefs or opinions as differentiating the personalities of the 
sexes? Unless the items used in this experiment represent an 
unfortunate selection it would appear that sex differences of the 
kind in question have probably been much exaggerated; that 
they are perhaps almost negligible. It is of course conceivable 
that a much better selection of trial items could have been made 
by a more penetrating observer of people. The result cannot, 
however, be attributed to lack of industry in assembling items 



EXTENSION OF THE M-F TEST 45 

for consideration or to lack of attempt to apply critical judgment 
in selecting from the larger number those most worthy of trial. 
Each item was selected either on the basis of its relationship to 
known data or on the basis of a definite hypothesis. The trial 
list is reproduced at the end of this section in order that the 
reader may judge for himself whether the relative failure of 
the experiment can be justly blamed upon the selection of items. 
The few which discriminated appreciably between sex groups are 
indicated by an asterisk. 

A few illustrations may be given of the kind of reasoning that 
went into the construction of the test. Statements like 2, 18, 
29, 36, S3, 57, 63, 78, 84, and 95 were selected on the theory that 
women tend to be more religious than men and to be somewhat 
more orthodox in their religious beliefs. Statements 12, 23, 33, 
39, 54, 59, 60, 68, 73, and 85 were expected to be more acceptable 
to men because of the element of aggression or will to power 
which they involve. Item 8 was chosen because women are 
supposed to be more conservative about divorce than men; 
items 14, 16, and 30 because investigations have shown them to be 
more superstitious; items 4, 5, 13, 28, 34, 35, and several others 
because women are thought to be more tender and sympathetic; 
item 43 because they are often said to be more given to "white 
lies"; item 34 because Lincoln symbolizes feminine sympathy 
and Washington masculine strength. And similarly throughout 
the test. The frequency with which seemingly good hunches 
were belied by the response statistics inclines us to believe that 
men and women do not differ from each other very greatly in the 
opinions they hold about commonly discussed issues. 

*1. The face shows how intelligent a person is. 
2. Laws should be passed to compel people to observe the Sabbath. 
*3. Blondes are less trustworthy than brunettes. 
*4. The weak deserve more love than the strong. 
*S. The hanging of murderers is justifiable. 

6. A square jaw indicates will power. 

7. It is better to be poor in the city than rich in the country. 

8. Nine-tenths of the divorces granted are unjustified. 

9. Character is indicated by the shape of the head. 



46 SEX AND PERSONALITY 

*10. The world was created in six days of twenty-four hours each. 

*11. Married women ought not to be permitted to teach school. 

12. Will power is supreme. 

13. It is justifiable to whip a baulky horse. 

*14. One usually knows when stared at from behind. 
15. Silent men are good thinkers. 

*16. Lines in the hand foretell the future. 

17. The "insanity plea" as a defense in murder trials should be 

abolished. 
*18. Preachers have better characters than most persons. 

19. The world is rapidly getting better. 
*20. Men are created equal in mental capacity. 

21. Conscience is an infallible guide. 

22. It is silly to let others see your emotions. 

23. Individual liberty is more to be desired than wealth. 

24. A wicked person cannot be happy. 

25. The pen is mightier than the sword. 

*26. Love "at first sight" is usually the truest love. 
*27. Inventors deserve more honor than artists. 

28. Criminals are really sick and should be treated like sick persons. 

29. God is a real personality who thinks and feels and wills. 

30. The stars have an influence on character. 

*31. Girls are naturally more innocent than boys. 
*32. An ugly face usually goes with a kind heart. 
33. A person is justified in taking another's life to save his own. 
*34. Lincoln was greater than Washington. 
*35. Hunting and fishing are wrong because cruel. 

36. In Bible times miracles actually occurred. 

37. Human beings are descended from the lower animals. 

38. Brains and beauty do not go together. 

*39. The United States should adopt a more aggressive foreign policy. 
*40. There is plenty of proof that life continues after death. 

*41. The largest fortunes should be seized by the government and 
divided among the poor. 

42. All human lives are equally sacred. 

43. "White lies" are sometimes justifiable. 
*44. Opportunity knocks but once for any man. 



EXTENSION OF THE M-F TEST 47 

45. What we call conscience is mainly the fear of being caught. 

*46. Children should be taught never to fight. 

47. Immorality is as wicked in men as in women. 

48. Paupers are usually themselves to blame for their condition. 

49. A shifty eye indicates dishonesty. 

50. Children should sometimes be punished by whipping. 

51. Rich men give to charity mainly to advertise themselves. 
*52. It is more important to be just than to believe in God. 

53. The clergyman's profession is nobler than the physician's. 

54. God is more a being of power than of tenderness. 

*55. It is better to teD your troubles to your friends than to keep them 
to yourself. 

56. Only those of average or superior intelligence should be allowed 

to vote. 

57. The wicked will suffer eternal punishment in the life to come. 

58. Students these days are not required to study hard enough. 

59. We are sometimes justified in returning evil for evil. 

*60. Wealth, power, and honor usually go to those who deserve them. 

61. Suicide is never justifiable. 

62. A receding chin means a weak character. 

63. Faith can heal a broken bone. 

*64. There should be perfect equality between men and women in all 
things. 

65. To be well liked while living is better than eternal fame after 

death. 

66. The artist is more to be admired than the statesman. 

67. The humblest laborer should be paid as much as the good mechanic 

because his family needs as much. 

68. The soldier is more to be admired than the musician. 

69. Riches should be totally disregarded in choosing a husband or 

wife. 

70. We should never send a man to prison for his first offense. 

*71. Green-eyed people are not to be trusted. 
*72. Women are purer than men by nature. 

73. We are sometimes justified in refusing to forgive our enemies. 

74. "Our nation, right or wrong," is a good slogan. 

75. Success depends more on luck than on ability. 



48 SEX AND PERSONALITY 

76. It can be proved that the soul is immortal. 

77. If a man becomes a criminal it is usually the fault of society. 

78. The religious missionary is a meddler and nuisance. 

79. Handsome men are less dependable than homely men. 

80. No one is entirely unselfish. 

81. Exactly the same standard of morality should apply to men as to 

women. 

82. It is better to live a coward than to die a hero. 

83. Unclean bodies are as bad as unclean thoughts. 

84. When science and the Bible disagree we should always believe the 

Bible. 

85. The United States should take possession of Mexico and civilize 

that country. 

86. It is possible even today to live exactly as Christ would have us. 

87. No one ever truly loves but once. 

88. There is no real justice in the United States. 

89. Killing animals for meat is morally objectionable. 
*90. We should never give to beggars. 

91. An illegitimate child should have the rights of inheritance. 

92. War is certain to be abolished sometime. 

93. The morals of the young are growing worse. 

94. If a man is moral it doesn't matter whether he is religious. 

95. The laws of nature are sometimes suspended in answer to prayer. 

96. The unmarried mother deserves the scorn she gets. 

TEST OF INTROVERTIVE RESPONSE 

An experiment was initiated in 1926 for the purpose of exploring 
sex differences in introvertive tendency. Cady's modification 
of the original Woodworth personal data sheet 1 served as a point 
of departure. As the Cady test had been administered a few 
years previously to the Stanford group of gifted children (IQ 140 
and above) and to a control group of unselected children, this 
offered an opportunity to discover whether any considerable 
number of its 85 items were promising for inclusion in the M-F 
test. Item counts were therefore made for 100 boys and 100 

1 CADY, V. M., The estimation of juvenile incorrigibility, /. Delinq. Monog., 2, 
Whitticr (Calif.) State School, 1923, pp. 140. 



EXTENSION OF THE M-F TEST 49 

girls, between the ages of 12 and 15. Half belonged to the 
gifted and half to the control group. Of the 85 items, 51 yielded 
sex differences large enough to warrant further trial. Only 
those were retained which gave critical ratios of 2 or higher for 
either the yes or the no response. The wording of many of the 
items was changed before printing for tryout. These 51 items 
will be referred to as Introversion, Series 1. 

Introversion, Series 2, was prepared concurrently with Series 1. 
It consisted of 47 items, about half of which were borrowed 
either from Laird's Cl and C2 or from the Heidbreder 1 list, the 
latter in advance of publication. The remainder of the items 
were new. 

Series 1 was given to 261 subjects in the seventh grade and to 
210 who were in the first year of high school. Series 2 was given 
to 231 subjects in the seventh grade and to 231 in high school. 
All the groups were composed of boys and girls in approximately 
equal numbers. On the basis of item counts for both series, 
scoring keys were made for both weighted and unit scores. As 
in all the other experiments, weighted scores showed so little 
superiority over the unit scores that they were discarded. 

The average reliability of Series 1 for four single-sex groups of 
100 subjects each was .44, and for a combined-sex group of 200 
subjects, .59. The sex overlap for seventh-grade and high- 
school subjects combined was 28 per cent. The average relia- 
bility of Series 2 for eight single-sex groups of 50 subjects each 
was .24, and for all subjects, sexes combined, .43. 

The 84 items composing Exercise 7 of the present M-F test 
(42 in each form) were selected as the best from Series 1 and 
Series 2. The 14 discarded were items which did not show a 
significant sex difference in both the yes and the no response. 
The retained items of both Series 1 and Series 2 were equally 
divided between Form A and Form B of Exercise 7. 

As was to be expected from previous data, the reliability of 
Exercise 7 is very low, averaging only .24 for eight single-sex 

1 HEIDBREDER, EDNA F., Introversion and extroversion in men and women, 
/. Abn. and Soc. Psychol., 1927, 22, 243-258. The authors are indebted to 
Dr. Heidbreder for supplying a manuscript copy of her test. 



SO SEX AND PERSONALITY 

groups. A sex overlap of 25 per cent was found for seventh- 
grade subjects, of 33 per cent for high-school subjects, and of 
34 per cent for college subjects. Reliability of the total score of 
Form A and Form B combined is only in the vicinity of .40 for a 
single sex, so that there is no warrant for the use of Exercise 7 
in the comparison of individual subjects. Its use, with due 
caution, is permissible only in the comparison of rather large 
populations. 

As in the case of the test of opinions, the question arises 
whether the results of this experiment should have been discarded 
altogether. The answer would have to be in the affirmative 
if the purpose of the M-F battery had been to provide a basis for 
profile ratings of individual subjects in a number of aspects of 
human personality. Several measures of sufficient reliability 
to be used in this way are unquestionably a desirable objective 
for future work, but the goal of the present series of experiments 
has been primarily the more modest one of sampling a wide 
field of sex differences the significance of which would be reflected 
in a single total score for the battery. Exercise 7 has, therefore, 
been retained for whatever it may add to the general picture. 

In the period that has elapsed since the M-F test was devised 
many studies have been made of the aspect of personality 
commonly designated by the term introversion-extroversion, and 
several scales for its measurement have been devised. The 
concept is one which varies greatly from author to author, with 
the result that some of the tests designed to measure the trait in 
question yield very low correlations with some of the others. 
However, three of the best known (Laird's, Thurstone's, and 
Bernreuter's) yield intercorrelations which approximate unity 
when corrected for attenuation, though whether the aspect of 
personality measured by these three tests has been suitably 
named is a question with which we need not here concern our- 
selves. Exercise 7 closely resembles in content the three tests 
just mentioned. Its low reliability, especially as compared with 
the Thurstone and Bernreuter tests, is in part attributable 
to the smaller number of items, but chiefly to the fact that it is 
here scored as a sex difference test instead of as a direct measure 



EXTENSION OF THE M-F TEST 51 

of introversion-extroversion. Its use in this way is justified not 
only by the data presented in this section but also by the sex 
differences which several later investigations have brought to 
light. These are unanimous in showing females somewhat more 
introverted than male subjects of comparable populations. The 
fact that the sex difference is not large probably has some 
connection with the low reliability of the test when used as a 
measure of masculinity-femininity. 1 

1 The reader who is interested in the field of psychological sex differences 
drawn upon in the construction of the seven exercises constituting the M-F 
Test is referred to a summary of the recent literature with a bibliography of 
older summaries by Catharine C. Miles, Sex in Social Psychology, Chapter 16, 
pp. 683-797, in A Handbook of Social Psychology (Ed. C. Murchison) 
Worcester, Mass.: Clark Univ. Press, 1936. Pp. 1195. 



CHAPTER IV 
THE M-F TEST AS A WHOLE 

The preceding chapter has given a brief and rather general 
account of the experiments which resulted in the various " exer- 
cises" or parts of the M-F test. It remains to present certain 
data and to discuss a number of issues which concern the test 
as a whole, including the weighting of items, method of deriving 
a total score, intercorrelation of the subtests, reliability, validity, 
establishment of norms, equivalence of Form A and Form B, 
correlations with other measures, sample score distributions, and 
desirable lines of extension and revision. 

THE WEIGHTING OF ITEMS 

The question of chief concern in this connection is whether 
the plan adopted whereby each item is given a unit score (M or F)- 
is as satisfactory as giving the item a score weight in proportion 
to its efficiency in discriminating between comparable sex groups. 
If it is not, then no ordinary degree of economy of time in scoring 
the test and in dealing with the measures it yields would justify 
our choice of the simpler method. By what criteria shall the 
relative merits of the two methods be judged? We are aware 
of only three: (1) reliability, (2) validity as indicated by per cent 
of overlap of score distributions for comparable male and female 
groups, and (3) validity as indicated by correlations with inde- 
pendent measures or ratings of masculinity and femininity. 

The third criterion should theoretically be the one most 
nearly crucial, especially if it includes correlations with clinical 
case histories as well as correlations with more strictly quantita- 
tive data. Unfortunately, it is a method we have not been able 
to use. There is available no other psychometric technique for 
measuring M-F differences against which the present test can be 
validated, and clinical data in sufficient quantity have not yet 

52 



THE M-F TEST AS A WHOLE S3 

been accumulated. There remains the outside criterion of 
M-F ratings of subjects by presumably competent judges, the 
difficulty here being that we have not been able to find two 
observers whose ratings of the same subjects agree to more than 
a trifling extent. The choice between unit scores and weighted 
scores has accordingly been determined primarily by the effect 
upon reliability and sex overlap and secondarily by considerations 
of convenience and economy. 

When the experiments in M-F testing were undertaken it was 
assumed that weighted scores would probably be found both 
more reliable and more discriminative. Their desirability was 
questioned only on certain theoretical grounds; whether, for 
example, the decrease in sex overlap which weighted scores were 
expected to bring about is necessarily a desirable result from 
the psychological point of view. No one knows what amount of 
overlap would be found if the totality of M-F traits were ade- 
quately and reliably measured. It is conceivable that the use of 
weighted scores might result in a spuriously small amount of 
overlap and thus exaggerate the sex differences which actually 
exist. It can be plausibly argued that heavy weights for indi- 
vidual items place a premium on sex differences that are acci- 
dental rather than fundamental accidental in the sense that 
they reflect almost entirely sex differences in experience. There 
are doubtless a few items of information which practically every 
woman but only a few men could answer correctly which would 
have a high discriminative index but which would have no 
psychological significance in the present study, and similarly, 
items of specifically male knowledge could be propounded. 
Items of this kind tell us nothing beyond the fact that the sub- 
ject is male or female, which we already know. The principle 
involved here does, as a matter of fact, apply in greater or less 
degree to every item in the M-F test. One may reasonably argue 
that the only way to avoid making the test too much a measure 
of accidental differences in experience is to take account only 
of the number of M-F differences without regard to their size. 

It was this point of view which led us in the first place to 
compare weighted and unit scores for reliability and validity, 



54 SEX AND PERSONALITY 

though it was hardly hoped that the data in these criteria would 
warrant the adoption of the method which we regarded as 
otherwise theoretically preferable. As it turned out, more than 
a hundred reliability coefficients computed by the two methods 
showed an average difference of less than .05 in favor of weighted 
scores. Even more unexpectedly, the average difference in sex 
overlap was less than 2 per cent. Although the average differ- 
ence both for reliability and for overlap favored the weighted 
scores, it was so small as hardly to justify the retention of a 
method which was both clumsy and psychologically question- 
able. Unit scores are therefore used throughout the test. Each 
response of the subject is scored + or , that is, masculine or 
feminine. The total score of a given exercise is the algebraic 
sum of the + and scores. It will be noted that in Exercises 2, 
3, 6, and one section of 5 omissions as well as positive responses 
are scored. 

The weighted item scores used for a time and then discarded 
in favor of unit scores were derived by the formula: 1 



where 

ad be 



This formula takes account of the fact that the reliability of the 
sex differences with respect to relative frequency of a given 
response is lower when the frequencies are small than when they 
are large. 

WEIGHTING OF EXERCISES FOR TOTAL SCORE 

The weighting of subtests in a battery of intelligence or special 
ability tests has become a fairly standardized procedure. Every- 
thing else equal, those subtests are assigned the most weight 
which have highest reliability, which correlate best with inde- 
pendent measures of the ability, and which duplicate least the 

1 COWDERY, K. M., op. cU. 



TEE M-F TEST AS A WHOLE 



.51 



other subtests of the battery. While agreement is general that 
these are the important factors to be taken into account, the 
precise weight to be given to each is a matter on which experts 
sometimes differ. Over and above these three factors is the 
problem of equating the several subtests for variability. This, 
however, is a question of wwweighting rather than weighting, for 
it is clear that a given subtest A yielding score distributions with 
twice as large standard deviations as those of subtest B carries 
automatically twice the weight of the latter. In order to give 
the two subtests a truly equal weight, the score on B must be 
multiplied by 2. 

The weighting of subtests in a battery of the type with which 
we are here concerned should take account of reliability, sex 
overlap (validity), intercorrelations, and dispersion of scores. 
The following data are pertinent: 



The subtests 


1 


2 


3 


4 


5 


6 


7 


Approximate average reliability, ten 
single-sex groups 


.40 


25 


.50 


.89 


.60 


.54 


.24 


Approximate average reliability, 
sexes combined 


"762' 


.34 


68 


(^fT 


.80 


64 


32 


Average per cent of overlap 


17 


30 


17 


f v 

26 


10 


31 


31 


Average correlation with the other 
subtests single-sex groups. . . . 


.01 


.07 


19 


15 


10 


09 


.13 


Average correlation with the other 
subtests sexes combined . 


.40 


.30 


47 


39 


.49 


27 


.38 


Approximate average of S.D.'s of 
16 groups .... 


7 


1 


8 


20 


30 


7 


2 



















The intercorrelations of the exercises may be ignored, as they 
are all relatively low. As for reliabilities, those of Exercises 4 
and 5 are highest, those of 2 and 7 very low, and the others 
medium. Overlap is least for 5, greatest for 2, 4, 6, and 7, and 
moderate for 1 and 3. The standard deviations of scores are 
large for 4 and 5, very small for 2 and 7, and intermediate for 
1, 3, and 6. The weights we have assigned are as follows, the 
figures given being the multipliers to apply to the raw scores 
of the several exercises before adding for total score: 



56 SEX AND PERSONALITY 

Exercise 12 3456 7 

Multiplier 1 % 1 1 2 1 H 

Perhaps no two psychologists on the basis of the available data 
would have arrived at exactly the same weights; certainly it 
would not be easy to defend those assigned as the best possible. 
As a matter of fact, within reasonable limits the weights assigned 
to the separate exercises have little effect on the correlations 
which the total score will have with other variables. The low 
weights given Exercises 2 and 7 would seem to be justified by the 
low reliability and large overlap of these tests. It is possible 
that the weight assigned to Exercise 5 is higher than it should be 
as compared with that of 4 and especially as compared with those 
of 1, 3, and 6. However, the sex overlap for Exercise 5 is the 
lowest found for any of the tests. 

EQUIVALENCE OF FORMS A AND B 

In apportioning the items between Form A and Form B con- 
siderable effort was made to insure that the two forms would be 
as nearly as possible equivalent both in respect to content and in 
respect to the mean scores which they would yield. Items were 
first laid out in pairs in such a way that the two items of a given 
pair would be psychologically as similar as possible. Account 
was next taken of the degree of the masculinity or femininity 
of each member of a pair, and pairings were juggled so as to 
arrive at a reasonable compromise between psychological and 
statistical equivalence of the forms. As more weight was given 
to psychological similarity of content, it turns out that scores 
on the two forms are not exact equivalents, those of Form B 
averaging a few points more masculine than Form A scores. 
The difference in mean total score on the two forms was 6.7 points 
for high-school boys, 10.2 points for high-school girls, 4.2 points 
for eighth-grade boys, and 3.4 for eighth-grade girls, all these 
differences being in the same direction. 1 Separate norms have 
accordingly been provided for the two forms. 

1 Groups to whom both forms of the test were administered took them half in 
the AB order and half in the BA order, although as it turned out the order of 
taking the forms did not seem to affect the score. 



THE M-F TEST AS A WHOLE 



57 



INTERCOKRELATIONS or THE EXERCISES 

Table 4 gives the intercorrelations of the seven exercises of 
Form A for two populations of SO each, the upper triangle for 
eighth-grade girls and the lower for high-school junior girls. 
The reliabilities, which are underlined, are in each case the 
average computed from several populations. Table 5 gives the 
algebraic averages of the correlations for the two groups. 

TABLE 4. INTERCORRELATION OP EXERCISES, FORM A 
(Upper right, eighth-grade girls; lower left, high-school girls) 



Ex. 


1 


2 


3 


4 


5 


6 


7 


1 
2 
3 
4 
5 
6 
7 


+ 40 


- 02 

+ 28 


+ .26 
+ 04 

+ .45 


-.21 
+ .24 

+ .35 
+ 89 


.00 

+ .05 
+ 18 
+ .03 
+ 60 


+ .11 
+ 04 
+ 46 
- 03 
- 01 
+ .54 


- 06 
-.08 
+ .17 
+ 20 
- 01 
+ 07 
+ 24 


+ .12 
+ .02 
-.08 
+ .14 
-.15 
- 01 


+ 07 
+ 20 
+ 06 
- 17 
+ 20 


+ 11 
00 
+ 47 
+ 17 


+ 26 
+ 21 
+ 52 


+ 05 
+ 40 


+ 02 





TABLE 5. AVERAGE OF CORRELATIONS FOR THE Two GROUPS OF TABLE 4 



Ex. 


2 


3 


4 


5 


6 


7 


1 


+ 05 


+ 14 


- 15 


+ 07 


- 02 


- 04 


2 




+ 06 


+ 22 


+ 06 


- 07 


+ 06 


3 






+ 23 


+ 09 


+ 47 


+ 17 


4 








+ 15 


+ .09 


+ 36 


5 










+ 02 


+ 20 


6 












+ 05 



It is evident that the several parts of the M-F test have little in 
common. In the preceding table of average intercorrelations 
only two values are found which can be regarded as clearly 
significant: (1) the correlation of .47 between Exercise 3 (informa- 
tion) and Exercise 6 (historical characters and opinions) ; (2) the 
correlation of .36 between Exercise 4 (emotional attitude) and 
Exercise 7 (introvertive response). The intercorrelations would 
still be very low even if they were corrected for attenuation. 



58 SEX AND PERSONALITY 

The two exercises which have highest reliability (4 and 5) yield 
an average intercorrelation of only +.15 and +.10, respectively, 
for the two populations. 

The intercorrelations would be slightly higher with populations 
less homogeneous than those of Tables 4 and 5. The inter- 
correlations of the exercises are higher when computed for the 
sexes combined, but intercorrelations thus derived have doubtful 
significance. 

The only conclusion possible from the data at hand is that 
M-F differences are so largely specific for the various types of 
items composing the separate exercises that search for a general 
factor or group factors by the application of factor-analysis 
techniques would be futile. It is doubtful in any case what, if 
any, psychological meaning attaches to the alleged factors 
disclosed by such techniques. Instead of adding to the con- 
troversial literature on this question we have employed (in 
Chapter XVI) a method of analysis which we regard as psy- 
chologically more meaningful. 

The total M-F score, as the test is now constituted, is a 
composite of samplings from several areas not highly correlated 
with each other. Warrant for its use rests on the hypothesis 
that a high average of masculinity or femininity in the fields 
covered by the test probably affects the total personality picture 
and has significant correlates in everyday behavior. The 
correctness of this hypothesis has not been conclusively demon- 
strated but is supported by a certain amount of clinical and other 
data. It is desirable at present to keep an open mind, for it is 
conceivable that behavioral correlates of appreciable amount 
will turn out to be for the most part specific to particular 
M-F fields, such as information, emotional attitudes, interests, 
etc. The specificity may carry even further; for example, to 
anger, fear, pity, etc., in the field of emotional attitudes, and to 
occupations, people, books, etc., in the field of interests. 

Perhaps the next step in M-F research should be to devise 
highly reliable subtests in as many fields of sex differences as 
possible so as to make profile studies of individual subjects 
feasible. The requisite reliabilities could doubtless be obtained 



THE M-F TEST AS A WHOLE 59 

for at least one form of each of the following subtests: word 
association, information, anger response, fear response, pity 
response, disgust response, ethical response, and interests. The 
sacrifice of duplicate forms would not be a serious loss, as these 
are not required for the types of research and clinical practice 
which are at present most needed. Pending such revision, 
provisional profile studies should be carried out with three or 
four of the present subtests by combining the scores of Forms A 
and B to improve reliability. 

RELIABILITY 

The reliability of total score of the M-F test has been com- 
puted both by the split-half method and by correlating Form A 
with Form B. The average of six reliabilities by the split-half 
method, for single-sex populations ranging from SO to 170, was 
.78, and for three combined-sex groups, .92. The corresponding 
correlations of Form A against Form B were .72 and .90. The 
reliability of the composite score of Form A and Form B is .88 
by the split-half method for single-sex groups, and .96 for 
mixed-sex groups. The probable error of an individual's score in 
M-F points is for one form of the test approximately 15 score 
points, roughly a third of the standard deviation of the score 
distribution of a typical single-sex group. If both forms of the 
test are administered the average score of the two has a probable 
error slightly in excess of 10 points, or about 22 per cent of the 
standard deviation of a typical distribution. It is of course 
extremely important not to lose sight of the measure's probable 
error in comparing the scores of individual subjects. 

Split-half reliabilities of the seven exercises have been computed 
for seven to ten different narrow-range populations with N's 
ranging from 50 to 100 each (eighth-grade, high-school, and 
college students). The averages of these reliabilities for the 
several exercises are given in Table 6. 

The total score of one form of the test is reliable enough (.92 
for sexes combined) to determine a subject's status fairly accu- 
rately in the total M-F range. The combined score of the two 
forms of the test accomplishes this with exceptional accuracy 



60 



SEX AND PERSONALITY 



TABLE 6. RELIABILITIES OF THE SEPARATE EXERCISES 





Single-sex groups 


Sexes combined 


T? * A 








No. of groups 


Av. rel. 


No. of groups 


Av. rel. 


1 


5 


40 


3 


62 


2 


7 


25 


2 


34 


3 


10 


50 


2 


68 


4 


7 


89 


2 


90 


5 


8 


.60 


2 


80 


6 


7 


54 


2 


64 


7 


8 


.24 


2 


32 



(reliability .96). The reliabilities of the seven parts of the test 
taken separately vary greatly from .24 to .89 for single-sex 
groups and from .32 to .90 for sexes combined. Only Exercise 4, 
emotional attitudes, is reliable enough to locate an individual 
subject with reasonable accuracy if only one form of the test 
has been given. If both forms have been given, all the other 
exercises except two are reliable enough to be useful in com- 
parisons of moderately small populations, but not reliable 
enough for comparing individual subjects. The two exceptions 
are Exercise 2, blots, and Exercise 7, introvertive response, both 
of which are so unreliable that they can be used separately only 
in comparing extremely large populations. Accordingly, if 
profiles are made of individual subjects these by rights should be 
based upon the combined score of Form A and Form B, and 
Exercises 2 and 7 should even then be omitted from the profile. 

As we see it now, it would perhaps have been preferable to have 
included in the M-F test only such subtests as could have been 
made highly reliable, even if only a single form could have been 
provided. Exercise 4, emotional attitudes, already satisfies this 
standard. Exercises 2 and 7 (blots and introvertive response) 
could not be made to do so without lengthening them to an extent 
hardly practicable. Exercises 1 (word association), 3 (informa- 
tion), and 5 (interests) would satisfy the standard well if provided 
with 50 per cent to 100 per cent more items than are included in 
both forms as they stand at present. 



THE M-F TEST AS A WHOLE 61 

TYPICAL CORRELATIONS WITH OTHER VARIABLES 

We have brought together in Table 7 some of the more interest- 
ing correlation coefficients found for M-F scores and other 
measures. Most of the relationships which enter into this table 
have been treated at length in other chapters, especially those 
with age, education, intelligence, interests, occupation, family 
background, and physical measurements. At this point we 
should call the reader's attention to the following outstanding 
results of the correlational work: 

1. Nearly all the correlations are low. 

2. Correlations of total M-F score with age, within the rela- 
tively narrow range represented by the populations in question, 
are so low as to be merely suggestive of a slight positive relation- 
ship (though we shall see later that over wide age ranges the 
correlation is significantly negative for males). 

3. Correlations of total score with intelligence approximate 
zero for males and tend to be slightly positive for females. 

4. Total score shows a significant positive correlation with 
Stenquist mechanical-ability test for males but not for females, 
and a lower but possibly significant correlation with McQuarrie 
mechanical-ability scores. 

5. M-F total correlates negatively with both the Laird and the 
Conklin tests of introversion, but not with the Neyman-Kohlstedt 
test. 

6. Correlations with Conklin introversion, Allport ascendance- 
submission, Pressey X-0, and Watson fair-mindedness differ as 
between college men of high scholarship and low scholarship, 
Conklin introversion being more negative and Allport ascendance 
more positive with the high-scholarship group, and the Pressey 
X-O more negative with the low group. 

7. College women who engage in many extracurricular activi- 
ties do not tend to test especially masculine. 

8. M-F score shows no appreciable correlation with the total 
score (or with any of the seven parts) of the Raubenheimer-Cady 
battery of character tests. 



62 SEX AND PERSONALITY 

TABLE 7. CORRELATIONS OF M-F SCORES WITH OTHER VARIABLES 



Variables correlated 


Population 


Sex 


N 


r 


P.E. 
of r 


Total M-P and age 


Eighth grade 
Eighth grade 
H.S. juniors 
H.S. juniors 
Eighth grade and juniors 


M 
F 
M 

F 

M 


95 
84 
79 
86 
174 


03 
.28 
-.03 
-.08 
14 


07 
07 
08 
07 
.06 




Eighth grade and juniors 
Stanford students 
Gifted 
Gifted 


F 
P 
M 
P 


170 
91 
79 
81 


.16 
- 03 
.08 
20 


05 
08 
08 
07 


Total M-P and Thorndike intel. . . . 


Stanford students 
Stanford students 


M 
P 


97 
92 


.00 
16 


.07 
07 


Total M-F and Terman group M.A 


Eighth grade 
Eighth grade 
H.S. juniors 
H.S. juniors 
Student nurses 


M 
F 
M 
P 
F 


95 
79 
78 
85 
74 


.13 
-.01 
.03 
.27 
20 


07 
.08 
.08 
.07 
07 


Total M-P and Scholarship av 


Stanford students 


P 


89 


- 12 


07 


Total M-F and Binet IQ 


Gifted 


M 


79 


06 


.08 




Gifted 


F 


81 


05 


08 


Total M-P and Extra-curric. act. . 


Stanford students 


K 


90 


- 09 


07 


Total M-F and Stenquist 


H.S. juniors 


M 


64 


.24 


.08 




H.S. juniors 


P 


64 


06 


08 


Total M-F and Laird C2 


Stanford students 


P 


90 


- 29 


09 


Total M-F and total of 7 Rauben- 
heimer character tests . . 


Older delinquents 


M 


329 


06 


.04 


Total M-F. and Neyman-Kohlstedt 
introversion 


Older delinquents 


M 


329 


02 


04 


Total M-F and Stenquist . . 


Older delinquents 


M 


329 


30 


04 


Total M-F and McQuame mechan- 
ical ability 


Older delinquents 


M 


329 


13 


04 


Total M-F and Conklin 


High-scholarship college 


M 


"46 


- 52 


07 




Low-scholarship college 


M 


46 


- 24 


09 


Total M-F and Allport A & S 


High-scholarship college 


M 


46 


36 


09 




Low-scholarship college 


M 


46 


08 


10 


Total M-F and Pressey X-O 


High-scholarship college 


M 


46 


03 


10 




Low-scholarship college 


M 


46 


- 44 


08 


Total M-F and Watson fair-minded- 
ness 


High-scholarship college 


M 


46 


18 


.10 




Low-scholarship college 


M 


46 


- 17 


.10 


Total M-F and 
girth of 9th rib (expanded) X 100 




P 


82 


26 


07 


stature 


Stanford students 


F 


70 


04 


08 


Total M-P and height 


Stanford students 


M 


32 


33 


11 


Total M P and sitting height X 10 


d ts 


M 


35 


30 


10 


*" stature 












Exercise 4 and Terman group M.A. 


Student nurses 
H.S. juniors 


P 
P 


74 

85 


38 
31 


0* 
07 


Exercise 4 and Terman group IQ 


H.S. juniors 


P 


85 


28 


07 


Exercise 4 and Thorndike intel 


Stanford students 
Stanford students 


M 
P 


97 
92 


16 
36 


07 
06 


Exercise 4 and Laird C2 . 


Stanford students 


F 


55 


- 20 


09 


Exercise 5 and Thorndike intel 


Stanford students 


M 


97 


- 09 


07 


Exercise 5 and Terman group M.A. 


Student nurses 


P 


74 


- 08 


08 


Exercise 5 and Stenquist .... 


H.S. juniors 


M 


64 


.20 


.08 



THE M-F TEST AS A WHOLE 63 

9. There is suggested a small but significant correlation 
between M-F score and certain physical measures. 

10. Exercise 4 (emotional attitudes) correlates positively and 
to a significant degree with intelligence. 

It is interesting to examine the data of Table 7 for such bearing 
as they may have upon the question of validity of the M-F test. 
Among the correlations which may be interpreted as supporting 
its validity are those between M-F total and mechanical-ability 
scores, introversion, Allport ascendance, Pressey X-0, general 
intelligence of female subjects, and physical measurements 
(relationship here more doubtful). At least these findings are 
in line with beliefs that seem to prevail with regard to differences 
between the masculine and feminine extremes within either male 
or female groups. Certainly the opinion is commonly held that 
short men tend to be more effeminate and athletic women less 
ladylike than the average of their sex, that highly intellectual 
women are likely to be somewhat masculine, and that emo- 
tionality and introvertive or inferior attitudes are more char- 
acteristic of femininity than of masculinity. Data reviewed in 
later chapters tend on the whole to support our interpretation 
of the above evidence as indicative of validity. 

OVERLAP OF COMPARABLE SEX GROUPS 

One of the customary procedures used in validating a test of 
ability or of personality is to select two criterion groups of 
subjects on the basis of some outside measure and find out how 
well the two groups are differentiated by the test in question. In 
the case of ability tests this method of validation is readily 
applied because of the reliability with which such contrasting 
groups can be selected; the experimenter usually has little 
difficulty in making up groups which contrast greatly in general 
intelligence, artistic ability, musical ability, or mechanical 
ability. In the case of personality traits the task is much more 
difficult. Independently validated outside measures of such 
traits are not ordinarily available, and the nature of the traits 
is such that they cannot be reliably rated for a subject by his 
intimate acquaintances. For one thing, it is harder to define 



64 SEX AND PERSONALITY 

what the experimenter means by introversion, fair-mindedness, 
inferiority attitudes, social intelligence, etc., than it is to define 
what is meant by intelligence or musical ability. In the second 
place, the manifestations of personality differences are less 
clear-cut, less easily observable, than differences in ability. The 
result is that authors of nearly all kinds of personality tests 
have had difficulty in setting up criterion groups that could be 
known to represent extremes with reference to the personality 
trait in question. 

The same difficulty would have been experienced in devising 
the M-F test had it been necessary to select criterion groups 
on the basis of M-F ratings of subjects by their teachers or other 
intimate acquaintances. In fact M-F ratings, as we have shown 
by several experiments, seem to be less reliable than ratings of 
almost any other personality trait. Fortunately, however, we 
do not have to select our criterion groups in this way in order to 
validate the M-F test; the criterion groups are everywhere at 
hand boys and girls, men and women and the maleness or 
femaleness of either group is known with practically 100 per cent 
certainty. It is only necessary to administer the test to com- 
parable sex groups and note the amount of overlap in the dis- 
tributions for the two sexes. This has been done for several 
pairs of sex groups, the amount of overlap for each pair having 
been computed both for the total score and for the separate 
exercises. 

Before presenting the statistics on overlap of sex groups it may 
be well to point out that the figures given are valid only for the 
total score of the particular items which compose the test as a 
whole or the subtest in question. The very method of selecting 
these individual items insures that the overlap will be relatively 
small. Certainly if all possible items of the general type found 
in a given exercise had been included, the overlap would have 
been much greater than has been found. The smallness of 
overlap is simply an index of the extent to which the items 
included bring sex differences in response. It is not to be 
supposed that the sexes really differ from one another in their 
interests, attitudes, and thought trends as much as the small 



THE M-F TEST AS A WHOLE 65 

overlap on the M-F test and its exercises might at first suggest; 
the fact that the test is composed entirely of items selected 
on the basis of their M-F discrimination necessarily exaggerates 
the true differences. The excuse for using a method which has 
this effect is that the primary purpose of the test is to bring sex 
differences into relief by measuring the extent to which a sub- 
ject's responses diverge from the mean of his sex on just those 
test items to which the sexes do tend to respond differently. 

The reader will observe that the seven parts of the M-F test 
present considerable diversity with regard to amount of sex 
overlap in scores. This is partly a function of the length and 
reliability of the exercises, but it seems to be in part a function 
of their content. The relatively large overlap for Exercise 2 
(blots) and Exercise 7 (introvertive response) is largely accounted 
for by low reliability. On the other hand, Exercise 4 (emotional 
attitudes) has also a very large overlap notwithstanding the fact 
that it is highly reliable. The explanation may be connected 
with the fact that this exercise calls for a graded response instead 
of the all-or-none type. For this reason, and possibly for others, 
it would be a mistake to interpret the per cents of overlap as 
accurate indices of the relative validity of the exercises. Such 
would be the case only if all other factors were equal. Exercise 5 
(interests) stands out from all the others in the smallness of 
its overlap; it contains more items than any of the other exercises, 
holds second place for reliability, and contains a large proportion 
of items with a strong tendency to sex dichotomy of responses. 

The cautions that must be observed in interpreting overlap of 
the different exercises apply equally to the overlap figures for 
total score. These figures are valid only for a test composed 
of items identical with or very similar to those of the present 
M-F test. If, for example, we say that a given male subject 
tests as feminine as the average female, that a given female rates 
at the 30th percentile for males, etc., it will be understood that the 
comparison is based strictly on the test as it stands. If all 
possible likenesses and differences between the sexes were taken 
into account the relative status of the subjects compared might 
be considerably altered. 



66 



SEX AND PERSONALITY 



The expression "per cent of overlap" has been used with 
various meanings by various writers. Overlap as we have used 
the term is computed as follows? 

Cumulate (upwards) in terms of percentages the distribution 
having the higher mean. Cumulate downwards the distribution 
having the lower mean. Then the percentage found at the point 
in the two distributions where the cumulated percentages are 
equal is the index of overlapping. Interpolation is of course often 
necessary if one wishes to locate the points exactly. The 
percentages do not need to be cumulated much beyond the point 
where their values for the groups compared become equal. The 
following example will make the procedure clear. 



Score 
intervals 


Frequency 


Cumulative percentages 


Males 


Females 


Males 


Females 


20 


1 








19 











18 


3 








17 


7 








16 


12 








15 


16 








14 


23 








13 


31 




62.2 




12 


28 


1 


43.3 


.7 


11 


17 


1 


26 2 


1 4 


10 


13 


3 


15.9 


3.5 


9 


6 


7 


7 9 


8.5 


8 


3 


14 


4.3 


18 4 


7 


2 


23 


2.4 


34 8 


6 


1 


28 


1.2 


54 6 


5 


1 


26 


.6 


73 


4 




15 




83.7 


3 




9 






2 




7 






1 




4 











3 






N 


164 


141 







In the above illustration it is obvious that the percentages at 
the point of equality are approximately 8.2, hence the index of 



THE M-F TEST AS A WHOLE 



67 



overlap can be expressed as 8 per cent. By this method the 
overlap shown in Fig. 1 is the area AON (or AOM), and not 
(as it is sometimes expressed) the area AMN. 



M N 
FIG. 1. Illustration of the measure of overlapping. 

TABLE 8. PER CENT OF OVERLAP OF SEX DISTRIBUTIONS FOR Two 

POPULATIONS 
(N = 100 in each) 





Eighth 
grade 


Eleventh 
grade 


Average 


Exercise 1 
Form A 


23.33 


14.67 




Form B 


21.14 


15.11 


18.56 


Exercise 2 
Form A 


30.80 


36.63 




Form B 


29.80 


25.23 


30.61 


Exercise 3 
Form A 


12.00 


14.63 




Form B 


17.64 


23.60 


16.97 


Exercise 4 
Form A 


34.25 


21.43 




Form B 


27.14 


30.43 


28.31 


Exercise 5 
Form A ... 


6.00 


10.21 




Form B . . . 


12.29 


6.89 


8.84 


Exercise 6 
Form A. . . . . 


29.43 


29.59 




Form B . . 


38.00 


26.55 


30.89 


Exercise 7 
Form A 


32.57 


29.59 




Form B 


24.56 


33.49 


30.07 


Total score 
Form A 


8.71 


8.36 




Form B 


7.66 


7.21 


8.02 











68 SEX AND PERSONALITY 

CORRELATION or M-F SCORES WITH TEACHERS' ESTIMATES 

One of the usual methods of "validating" a mental test is to 
correlate its scores with ratings of the tested subjects by teachers 
or others. It is notorious, however, that the reliability of such 
ratings is usually low even for abilities (general intelligence, etc.) 
and that they are still lower in the case of personality traits. For 
general or special abilities the ratings of two supposedly compe- 
tent judges do not ordinarily correlate more than .4 to .6 and 
for personality traits from .2 to .4. Perhaps some would expect 
masculinity-femininity to be an exception to the rule in view of 
the confidence with which most of us judge our acquaintances 
on this personality trait. However, far from being an exception, 
M-F ratings yield particularly low reliabilities. The reasons 
for this have not been experimentally investigated, but probably 
lie in the varying concepts which people have in regard to what 
constitutes masculinity or femininity. Whatever the explana- 
tion, the limitations of ratings as a method of validating M-F 
scores are obvious. 

In one of our experiments 200 subjects of each sex enrolled in 
the eighth and tenth school grades were rated for masculinity- 
femininity by from one to six teachers who had had them in their 
classes for several months. The ratings were secured with the 
help of one of our former graduate students of psychology, Miss 
Lela Gillan, who taught in the school system where the experi- 
ment was made and who took over the entire task of instructing 
the raters and getting their fullest cooperation. The ratings were 
made, we believe, with more than ordinary care. Each teacher 
was given two lists of names of the subjects enrolled in her 
classes (one list for each sex) and was asked to take them home 
and at her leisure consider each pupil with respect to masculinity 
and femininity of behavior. Her task was to check in each list 
those she would rate among the one-fourth who were most 
masculine, and those she would rate among the one-fourth who 
were least masculine, leaving unmarked the middle SO per cent. 

The subjects were given one of the preliminary series of M-F 
tests including 91 information items, 170 interest items, 51 



THE M-F TEST AS A WHOLE 69 

introversion-extroversion items, and 60 word-association items. 
The correlation between composite teacher ratings and the M-F 
scores on these tests were +-10 for information, +.15 for interest, 
+.02 for introvertive response, and +.03 for word association. 
However, the reliability of the ratings was so low as to preclude 
any very considerable correlation with the test scores. The 
reliabilities were .142 for ratings of boys and .216 for ratings of 
girls by a single judge. These figures become .332 for ratings 
of boys and .453 for ratings of girls by three judges, and .498 for 
boys and .623 for girls when there are six judges. 

When the M-F test had reached its present stage one form of 
it (A) was given to 82 male Stanford students who, immediately 
after taking the test, were asked to rate themselves on degree 
of masculinity with respect to (a) childhood interests, (6) occu- 
pational interests, (c) present method of spending leisure time, (d) 
emotionality, and (e) general personality make-up. The ratings 
were made by the cross-on-a-line method and were preceded by a 
brief discussion designed to promote frankness of ratings. The 
subjects did not know the purpose of the test and it is believed that 
most of them were successfully misled to believe that the ratings 
had no connection with the test they had just taken. The five 
ratings correlated as follows with total M-F score: childhood 
interests, .08; vocational interests, .06; use of leisure, .22; emo- 
tionality, .21; general make-up, .13; average of the five self- 
ratings, .19 .07. 

A more careful experiment was carried out later to determine 
whether the correlations of ratings with M-F scores would be 
higher if judges were asked to rate, not masculinity or femininity 
in general, but more specifically defined traits which are pre- 
sumably related thereto. For each trait a composite rating was 
computed based upon the independent ratings of the two (or 
three) judges, and the correlation was found between this com- 
posite and the total score of the M-F test. Most of the corre- 
lations were in the expected direction, but only three of them 
were as large as three times their probable error. These were: 
with "tomboyishness," .33 + .08; with leadership,. 24 .07; 
with having manual interests typical of women, .25 .07. 



70 SEX AND PERSONALITY 

The fact that the M-F test is composed of items empirically 
selected as showing sex differences in responses makes it, ipso 
facto, a measure of mental masculinity and femininity in the 
scope embraced by its contents. One may therefore say that 
the test is inherently and of necessity valid. However, this 
kind of validity becomes a rather empty merit unless the scores 
yielded by the test can be shown to have demonstrable correlates 
in behavior. We believe that some of the data presented in 
later chapters strongly suggest the reality of such correlates, 
especially in the case of subjects who deviate from their sex 
norm to an extreme degree in either the masculine or the feminine 
direction. For validation of the test in this sense we shall have 
to look to clinical studies of such deviates rather than to person- 
ality ratings of the kind that are based upon relatively superficial 
and unreliable observations. The data on score distributions 
and score means in the following section constitute interesting 
evidence of validity. 

SCORE DISTRIBUTIONS FOR SELECTED GROUPS 

In Tables 9, 10, and 11 score distributions are given for a 
number of selected groups of both sexes. The data which they 
contain are fragments of material dealt with more fully in later 
chapters, but the reader may wish to examine a few distributions 
at this point both because of their general interest and because 
of their bearing upon the question of validity. In comparing 
the scores of the different groups the reader should consider 
whether the distributions and means are in line with what could 
be expected on the basis of common observation regarding the 
degree of masculinity or femininity of population classes of the 
kind in question. Is the small overlap of distributions for anal- 
ogous sex groups a general phenomenon or will it perhaps fail 
to appear with English youths, Americanized Japanese, negro 
college students, or delinquent boys and girls? Do athletes and 
professionally trained engineers, believed by many to be character- 
ized by masculine behavior, actually test more masculine than 
male groups commonly characterized as somewhat effeminate, 
such as artists, musicians, ministers, and theological students? 



THE M-F TEST AS A WHOLE 71 

Do passive male homosexuals score as feminine as they are 
reputed to be? Is the highly intellectual woman holding the 
M.D. or Ph.D. degree shown by the test to be more masculine- 
minded than the typical housewife or female domestic employee? 

Before examining the tables the following facts should be noted 
regarding the composition of some of the groups. The "general 
population" of Table 9 is fairly representative of the younger 
educated adults in California. The negroes were students in 
prominent negro colleges. The Japanese were school children 
in Hawaii and were almost all of the second generation of immi- 
grants. "Office workers/' both male and female, included 
office managers, secretaries, clerks, accountants, etc. The 
delinquents were inmates of two California state institutions, 
the Whittier school for boys and the Ventura school for girls; they 
are believed to be fairly representative of male and female delin- 
quents of the late teens who have received court commitments. 

The male athletes (Table 10) were mostly football players; 
the female athletes (Table 11) were somewhat less specialized. 
The engineers were university trained and were practicing their 
profession. The males classed as " teachers " were in many cases 
supervisors or principals, but the women so classed were nearly 
all actual teachers. The " gifted ' ' represented a random selection 
of the older of the 1,000 gifted children studied by the senior 
author; when given the M-F test they were nearly all between 
the ages of fifteen and twenty. The music groups were students 
in one of the best known schools of music in the country. The 
theological group was made up of three subgroups whose score 
distributions were very similar: students preparing to become 
Catholic priests, Protestant theological students, and Protestant 
ministers. The "artist" group was composed chiefly of pro- 
fessional painters in California. The homosexuals were nearly 
all male prostitutes, which means that the majority of them 
would probably classify as of the passive homosexual type. 

Table 9 illustrates a fact to which we have found no exception, 
namely, that markedly contrasting distributions are always found 
when comparable male and female groups are tested. We have 
no doubt that this will hold for all modern Occidental cultures. 



72 



SEX 4MD PERSONALITY 



I 

2 



grade 



Eigh 



r-t -( CS f-l 



W 

3 



High sc 



o o ~< 

^*^Of 

10 vo 

oo -44 

1 



O 

I 

s 



> OC r-> * 
O Cf\ t 



i 

U 



Os 

s 



populat 



o r 

^ ' 

1 



P. 






THE M-F TEST AS A WHOLE 



73 



CS Os u- OO t> C4 m 



> in O ^ 
cs ^ 

+ 



tfl 

^ 

1 



O\ *f) \r\ * 
00 10 



O\ xO C4 O 

**t SO < 

* * 



i 



74 



SEX AND PERSONALITY 



2 S 



< \o * < 

< <S O 

o ^ ^ 



I 

S 



g 







! 



^<M-iO\tSNOtOi-^O^S^< - 



r> oo O l o 



8 




1 i j I 1 l I l I l 1 



THE if-P TEST AS A WHOLE 



75 



II 



C4**r*r-.ou > >^'f > i - i 



ON to 

I 



ll 



~* O u-> O> 
10 O ON ON 

o ON NO 



I OO 10 
I CM to 



2 $ 

II 



O O IN. O 

10 - 10 sO 

^H O\ 10 

7 ^ 



I 



I 1-1 <-H OO 



8S 

NO i "^ 
I 



| 
O 



IN. ro 10 
to * 

I 



1 O O IN. 
10 (S OO 



q 





SO to 10 
10 10 00 

** oo O 



I I I I I I I I I I I 



CO CO 



76 



SEX AND PERSONALITY 



The difference between the means of comparable male and 
female groups is an expression of the test's validity. The 
smallest difference we have found that of 48 score points 
between female prostitutes and male homosexual prostitutes 
is 3.4 times its standard error, and the majority of differences 
are from three to four times as great. The following score 
differences between the sex means are typical : 



Group 


Score 
difference 
between 
means 


No. of 
males 


No. of 
females 


Eighth-grade children 
High-school juniors . . 


143 

155 


200 
196 


196 
180 


Gifted children (age 16-18) 
Delinquents . . . 


123 
132 


75 
129 


72 

54 


English private-school children 
Japanese adolescents. . . 


99 
100 


59 
32 


60 
41 


College sophomores 


128 


130 


130 


College athletes 


114 


29 


53 


College-of- music students. 
Negro college students . . 
Prostitutes 


87 
109 

48 


50 
51 

52 


50 
25 
12 


Adult population, age 20 to 29 
Adult population, age 50 to 59 
Adult population, age 70 to 85 


ni 

115 

89 


138 
82 
28 


285 
162 
94 



The above differences between means are nearly all between 
twice and three times the S.D. of the usual score distribution. 
Sexual inverts seem to be the only exception to the rule that 
males score far more masculine than females, and this exception 
is of course the best of evidence of the test's validity. Male 
inverts (passive male homosexuals) commonly score between 
+25 and 60, with a mean around 20. The small number of 
female inverts (active female homosexuals) we have tested have 
nearly all scored between +75 and 30. This reversal of the 
direction of sex difference in the case of inverts presents, in our 
opinion, one of the most challenging problems in the psychology 
of personality. 



THE M-F TEST AS A WHOLE 77 

DANGER OF SCORE-FAKING 

In any pencil-and-paper test of the questionnaire type that 
requires the subject to answer questions about how he thinks or 
feels, what he is interested in, or how he is accustomed to react 
in given situations, it is assumed that the subject will cooperate 
by responding as truthfully as he can. Failure thus to cooperate 
would probably invalidate the subject's score on almost any of 
the current personality tests. One would not need to be a 
psychologist to be able to score as fair-minded on the Watson 
test, extroverted on the Laird C2, or self-sufficient on the Bern- 
reuter inventory, provided one knew what the test was intended 
to measure. Unfortunately the authors of personality tests 
have ordinarily not taken the trouble to investigate the extent 
of this danger, although precautions are sometimes taken to 
keep the subject in the dark with respect to the purpose of the 
test. 

With the assistance of Dr. Lowell Kelly the M-F test was given 
three times to a group of 52 college sophomores (33 women and 
19 men) under the following conditions. (1) At the time of the 
first test the subjects were merely informed that the investigator 
was making a study of the interests and attitudes of college 
students; the real purpose of the test, we learned by later inquiry, 
was not suspected by a single one of them. (2) One week later 
they were again tested, but were first told what the test measures 
and were asked to show us how much they could influence their 
scores at will. More specifically, half of the men were requested 
to make their scores as masculine as possible, the other half to 
make them as feminine as possible. Similarly for the women. 
(3) A week later those who had been asked to make their scores 
masculine in the second test were now asked to make them as 
feminine as possible, and vice versa. 1 

The subjects were promised that they would be informed of their 
scores, and they entered into the experiment with great zest, as 

1 KELLY, E. LOWELL, CATHARINE Cox MILES, and LEWIS M. TERMAN, Ability 
to influence one's score on a pencil-and-paper test of personality, Char, and 
Personality, 1936, 4, 206-215. 



78 



SEX AND PERSONALITY 



though it were a kind of game. The results gave a very empha- 
tic answer to the question as to whether M-F scores can be faked. 
Following are the means and S.D.'s of the naive scores, the 
masculinized scores, and the feminized scores: 





Males 


Females 


M 


S.D. 


M 


S.D. 


Naive scores 


+ 66.8 
+208.8 
-140.5 


39 2 
68.9 
51.8 


- 56.2 
+ 189.1 
-147.6 


48.1 
93.9 
62.6 


Masculinized scores 


Feminized scores 





In other words, a typical group of males at the most masculine 
age are able to earn a mean score more feminine by far than the 
mean of any female group we have tested; a typical female 
group can make itself appear far more masculine than any male 
group we have tested! The distance between the masculinized 
and feminized means is 349.3 score points for the males and 336.7 
score points for the females. The shift amounts to seven or eight 
times the standard deviation of the usual naive score distribu- 
tion. Male subjects are able to shift their scores more in the 
feminine than in the masculine direction, females more in the 
masculine direction: the explanation doubtless lies in the fact 
that the test gives more room for reverse shifts. The ability to 
influence score was found to be uncorrelated with age, intelli- 
gence, or scholastic achievement within the ranges provided by 
this group. The ability should be further investigated as a 
possible index of "social intelligence" or "psychological insight." 

Since the faking of scores is so easy, it is fortunate that subjects 
almost never suspect the purpose of the test. Of the hundreds 
who have been asked, immediately after taking the test, what 
they thought we were trying to measure, so far not one has 
guessed correctly. As the test comes into more general use the 
possibility of score-faking will of course increase, but we do not 
believe that its validity for most purposes is ever likely to be 
jeopardized from this cause. The danger is present only when 



THE M-F TEST AS A WHOLE 79 

the subject has a definite motive in appearing other than he is. 
One of us gave the test to 31 college students (17 men and 14 
women) after explaining to them carefully what the test is 
intended to measure, yet the resulting mean for neither sex 
differed significantly from the college norms. The mean for 
the men was 3.6 points more masculine than the norm, that for 
the women S.I more masculine. 



CHAPTER V 

CORRELATIONS OF M-F SCORE WITH PHYSICAL 
MEASUREMENTS AND TRAIT RATINGS 

If masculinity and femininity are thought of in relation to 
health, robust would be popularly regarded as the masculine and 
frail as the feminine adjective. Furthermore, a correlation 
between M-F score and physical build should be expected if 
masculinity-femininity is an expression of personality type 
grounded in innate constitution. Popular belief rates the large 
woman as masculine, the small man as feminine; a man with 
large hips or a woman with broad shoulders is likely to be sus- 
pected of an anomalous constitutional trend. Fortunately we 
have been able to check these popular beliefs against physical 
data for sizable groups for whom M-F scores were available. 

M-F SCORE AND HEALTH RATINGS 

The M-F blank used in our later work calls for a self-rating 
of each subject with regard to health. The categories provided, 
one of which the subject is asked to check, are robust, above 
average, average , below average, frail. Table 12 gives the mean 
M-F scores of males and females of our adult population with 
high-school education in each of the five categories; also the 
means for the two highest categories combined and the three 
lowest combined. 

In view of the highly masculine scores of athletes, both male 
and female, and in view of the popular belief that frailty is asso- 
ciated with femininity, it is interesting to find that in this group 
of adult men those who rate themselves as robust or above 
average have a mean score which is 7 points, or .15 of a standard 
score, 1 lower than the mean of those in the three lower categories. 

1 See p. 477 ff. for table of standard scores. 

80 



CORRELATIONS OF M-F SCORE 



81 



TABLE 12. HEALTH AND M-F SCORE. HEALTH Is MEASURED IN TERMS OF 

SELF-RATINGS ON A FIVE-POINT SCALE 
Subjects: Adults of high-school education, ages twenty to sixty-five years 





No. of 


Mean 


Combined 


C T\ 


Diff. 




cases 


M-F score 


mean 


O.U.M 


S.D.diff. 


Males: 
Robust 


29 


+ 47 1} 








Above average 


67 


+ 44 5 '' 


+45 3 


5 68) 




Average 


87 


+ 51 4 






.96 


Below average. . . 


3 


+ 57 2 > 


+52 7 


5 22) 




Frail ... 


2 


+ 105 5 1 








Omissions 


24 


+ 58 8 




10 73 
















Total 


212 


+ 50 








Females : 
Robust 


63 


- 85 U 








Above average .... 


121 


- 76 6 } 


-79 5 


3 04) 




Average 


288 


- 89 ) 
i 






2 44 


Below average 


26 


- 89 5 > 

r 


-88 9 


2 38) 




Frail 


3 


- 76 2 \ 








Omissions. . . 


32 


- 78 6 




7 28 




Total 


533 


- 85 5 









The difference, however, is not significant, as it is just less than 
its standard error. We cannot, of course, assert on the basis 
of our findings that men of average or inferior health are more 
masculine than those who have superior health, but we can 
perhaps safely say that the converse is not true. The subjects 
who gave these self-ratings ranged in age from twenty to sixty-five 
years, and since frailty and femininity of M-F score both increase 
with age one might have expected the age factor itself to insure 
a positive correlation between good health and masculinity. 
The 24 men who did not rate their health averaged 13.5 M-F 
points more masculine than the men with superior health and 
6 M-F points more masculine than the group with average or 
inferior health, another bit of evidence, possibly, of the relation 
of indifference or noncooperation to masculinity. 

The results for women are just the reverse of those for men; 
women who rate themselves as average, below average, or frail 



82 SEX AND PERSONALITY 

in health have a more feminine score than those who rate them- 
selves as above average or robust. The difference is 2.44 times 
its standard error and therefore probably significant. Women 
who omit the rating, like men who do so, average more mascu- 
line than either of the other groups, but the difference is not 
reliable. 

Incidentally it is interesting to note that relatively twice as 
many women as men rate themselves as below average or frail: 
S per cent as against 2^ per cent. We are unable to say whether 
this represents a true sex difference in physique or whether it 
merely reflects the greater respectability of ill-health among 
women. It is unfortunate that the health ratings could not 
have been made by physicians on the basis of a thorough medical 
examination. It is possible that an investigation of this kind 
would disclose significant relationships between M-F score and 
special conditions of health, apart from any correlation that 
might be found between score and a generalized health rating. 

M-F SCORE AND PHYSICAL MEASUREMENTS 

It has been thought that a man being "a man to his finger 
tips" and a woman being "a woman even to her little toes," 
correlations should be demonstrable between physical measures 
and M-F scores. To test this view we have correlated M-F 
scores with (1) measures of physical size and (2) such measures 
of physical proportion as are recognized expressions of sex con- 
trast or difference. 

Series of measurements of 152 women and 138 men were avail- 
able for correlation. 1 Height, weight, and pulse were correlated 
directly with M-F score. Other measurements were trans- 
formed into ratios to stature so that the specific factor rather 

1 The women's measurements were available through the courtesy of Professor 
Helen Bunting of the Department of Physical Education for women, Stanford 
University. The measurements of the men were obtained with the kind per- 
mission of Dr. Thomas A. Storey of the Department of Physical Education for 
men. All besides the height and weight measurements for the men were made 
under the direction of Professor W. R. Miles of the Psychology Department in 
connection with another psychophysical research. 



CORRELATIONS OF M-F SCORE 



83 



than the general size element might be isolated for comparison 
with M-F score. 

TABLE 13. CORRELATIONS BETWEEN M-F SCORE AND PHYSICAL MEASUREMENTS 
IN Two GROUPS OF SUPERIOR COLLEGE WOMEN 



Physical measures 


Correlation 
with M-F score 
Group I 
(82 cases) 
Group II* 
(70 cases) 


Physical measures 


Correlation 
with M-F score 
Group I 
(82 cases) 
Group II* 
(70 cases) 


Pulse (before exercise) . 
Weight 
Height (standing) . . . 
Ratio to stature X 100 of: 


+ 152 .07 
+ 07 .07 
-.025 .07 


Ratio to stature X 
100 of: 
Girth of chest (ex- 
panded) ... 


+ 080 .07 


Height (sitting) 


046 .07 


Girth of chest (con- 








tracted) 


+ 075 .07 


Girth of waist 


+ .120 .07 






Depth of abdomen. . . 

Depth of chest (con- 
tracted) 


+ .174 07 
*-.084 .08 

+ .082 07 


Girth of 9th rib 
(expanded) . 
Girth of 9th rib 
(contracted) . 
Lung capacity 


+ 261 07 
*+.037 08 
+ 212 07 
*+ 033 08 
+ 185 07 


Breadth of shoulders 


+ .027 07 


Expansion (breadth 
of chest) 


* .055 .08 
-.065 .07 


Breadth of chest (ex- 
panded) 


+ 072 .07 


Expansion (girth of 




Breadth of chest 
(contracted). 


+ 138 .07 
*+ 002 08 


chest) 
Expansion (depth 
of chest) . . 


.086 .07 
-.191 .07 



* The starred coefficients refer to Group II. 

For a group of 82 women coefficients were derived for the 
correlation between score and (1) measurements of pulse, weight, 
and height, and (2) fifteen ratios to height as follows: sitting 
height, girth of waist, depth of abdomen, depth of chest (con- 
tracted), breadth of shoulders, breadth of chest (contracted), same 
(expanded), girth of chest (contracted), same (expanded), girth 
at ninth rib (expanded), same (contracted), lung capacity, 
expansion (breadth of chest). Of the 18 coefficients 2 were three 
times the P.E., S others were twice the P.E., and the remaining 
11 were less than 2 P.E. The two of possible statistical signifi- 
cance were between score and (1) the contracted and (2) the 



84 SEX AND PERSONALITY 

expanded girth at the ninth rib (indirectly vital capacity). In 
order to check the results for these two and certain of the other 
coefficients that deviated farthest from zero, correlations were 
derived for 70 other Stanford women for whom M-F scores were 
available. None of the coefficients in the second series was 
larger than its probable error, and the relationships that appeared 
largest in the first series were small in the second. One factor of 
difference between the two groups that has been suggested 
by later findings should, however, be mentioned. The first 
group included a considerable number of the best women athletes 
in college, as well as a general selection of college women. The 
second group included few women who were very superior in 
athletics. This difference in the composition of the groups might 
account for the regular and persistent discrepancy between 
the two series of coefficients and if so would leave us with some 
evidence of a tendency for M-F score to correlate with athletic 
physique. We have seen (Tables 10 and 11) that groups of 
college athletes, both male and female, do make outstandingly 
masculine scores on the M-F test. The matter deserves further 
investigation, but from the data at hand one cannot infer that 
there is any dependable relationship between M-F score and 
physical measurements in the case of women. The relationship 
vaguely hinted at in a single group is, if other than a chance 
occurrence, associated with the presence of an interest or behavior 
type already shown to correlate to some extent with M-F score. 
A group of 32 men of small stature (height range 160 to 179 
cm.) was examined as to the relation of physical traits and M-F 
score. Their average M-F score of +61.7 (s.s. of +.45) is lower 
than that of all but one of our groups of college men in their own 
or any similar institution. The one group that rates as low is 
a group of 46 men of exceptionally high scholarship (upper 
quartile). The men in the small stature group do not owe their 
placement to a like cause, as their average scholarship rating is 
just below the general average for the institution. Not only do 
these men as a group score low on the M-F scale, but the 12 
smallest of their number average even lower (+48) and the 
correlation between height and M-F score for the group is 



CORRELATIONS OF M-F SCORE 85 

+.33 + .ll,indicatingaprobablysignificantrelationship. Corre- 
lations for this group have also been worked out between M-F 
score and (1) weight, (2) ratio of weight to height, (3) ratio of 
height to weight, (4) ratio of shoulder to hip, and (5) ratio of 
hips to weight. Correlations were also made between score and 
shoulder width, and score and hip width. Of these correlations 
that with weight is two and a half times its P.E. The others 
are all nearer one than two times the P.E. Evidently an indi- 
cation of a correlation between an M-F index of build and the 
mental M-F score is not found in this group except as stature 
alone is such an index. 

Correlations between M-F scores and four ratios of body build 
have been derived from the data for a group of 106 college men 
in an elementary psychology class. These give the following 
values: 



r 



., . , .hip width X 100 _ , tl 

M-F score with ratio - . 20 .11 

stature 

_,_, shoulder width X 100 no 

M-F score with ratio . . .08 .11 

stature 

M-F score with ratio Wei f; X1 +.04+ .07 

stature ~ 

v u . sitting height X 100 _ , in 

M-F score with ratio 9 , . . . . 30 .10 

stature 

It may or may not be significant that one of these correlations 
(M-F score with ratio of sitting height to standing height) is 
three times, and another (that with ratio of hip width to stature) 
is twice, its probable error. Both of these correlations are in the 
direction to be expected if masculinity of build is associated 
with masculinity of score. It is well known that greater length 
of trunk in proportion to total height is one of the most charac- 
teristically feminine traits disclosed by physical measurements, 
and the correlation of .30 means that/0r this particular sample 
of the male population this factor is definitely associated with 
less masculine score. We can by no means be sure, however, 
that a larger and therefore more reliable sampling would yield 
the same result. 



86 SEX AND PERSONALITY 

At this point the reader may want to examine the data on 
physical measurements of 24 male homosexuals in Chapter XI. 
These subjects belonged to our most feminine-testing male group, 
but apart from stature, in which they were significantly below 
both unselected army recruits and college students, their physical 
measurements showed no significant departure from the male 
norms. No anomalies of the sexual organs were found. 

Summarizing, it is obvious that no definite conclusion can be 
ventured with regard to the existence or nonexistence of a rela- 
tionship between M-F score and physical measurements; the 
available data are based on small populations and contain con- 
tradictions. One can only say that if there is any correlation 
between score and any of the physical measures we have used 
it is probably not very large. The data available suggest that 
there may be for males a positive correlation between masculinity 
and height, and also between femininity and length of trunk in 
relation to height. In the case of females the only significant 
relationship suggested at present is the apparently positive corre- 
lation between score and ratio of girth at ninth rib (either 
expanded or contracted) to stature. The latter correlations for 
one group, partly composed of outstanding athletes, are for 
contracted girth three times the P.E., and for expanded girth 
nearly four times the P.E. A more searching investigation is 
urgently called for. Large populations should be measured and 
tested and if more than one measure yields an apparently reliable 
correlation the technique of multiple correlation could be applied. 
It is conceivable that the combination of a number of small but 
reliable correlations would yield a sizable multiple correlation 
with M-F score. Measures should also be made of amount and 
distribution of hair growth, of metabolism, and, when reliable 
techniques are available, of amount of gonadal hormones in the 
blood. 

M-F SCORE AND AGE OF PUBERTY 

The question was raised in an earlier chapter whether the 
course of M-F development is affected by early or late puberty. 
Such an influence might occur either as a direct effect of endocrine 



CORRELATIONS OF M-F SCORE 



87 



factors or as the effect of differential social treatment based 
upon recognized physical maturity or immaturity. 

It happens that records of approximate age of puberty are 
available for 80 boys and 96 girls who took the M-F test at the 
average age of 16}^, the range of ages falling between 13 and 19. 
The subjects were members of the California group of "gifted 
children" who had been under observation for several years. 
The record was the mother's statement as to the age when her 
daughter first menstruated or when her son's voice changed. 
Although records of this kind are known to involve appreciable 
error, especially in the case of voice change, they are probably 
accurate enough to reveal any close association that might exist 
between M-F score and age of sexual maturation. 

According to these records the ages at which puberty was 
attained were distributed as follows: 



Age 


10* 


11 


12 


13 


14 


15 


16 


17 


18 


19 


Bovs 






? 


11 


71 


w 


11 




7, 




Girls 


4 


IS 


?8 


?S 


10 


11 


? 






1 

























* Age 10 means 10 but not yet 11, etc. 

Each sex group has been divided into two categories by age, and 
biserial r's have been computed. Boys were classed as less than 
15, or 15 or over; girls as less than 13, or 13 or over. The result- 
ing correlations with M-F score were +.30 for boys and +.10 
for girls. 

Only the first of these correlations suggests a genuine relation- 
ship. However, in the case of both boys and girls a significant 
difference in M-F score is found where we limit the comparison 
to the extremes of the age distributions. For example, the 
13 boys whose voice changed before 14 had a mean M-F score of 
+65 ; the 13 whose voice changed at 16 or later, a mean of +44. 
The difference is about .5 S.D. of a typical male distribution. 
The mean M-F score of the 19 girls who menstruated before 1 2 
was 64; of the 24 who first menstruated at 14 or later, 33. 
The difference is approximately .75 S.D. of a typical female 
distribution. The data therefore rather strongly suggest that 



88 SEX AND PERSONALITY 

decidedly early, as compared with decidedly late, puberty is 
associated with greater masculinity in boys and greater femininity 
in girls. 

If this apparent influence of early and late puberty should be 
confirmed it would remain to discover whether the effect persists 
through adult life or disappears as maturity is reached; also to 
determine, if possible, whether the influence is primarily bio- 
chemical or mainly the result of social pressures. 

We have tried to estimate the possible effect of one type of 
social influence in the case of these subjects, by correlating M-F 
score with age-grade status in school. One might assume that 
marked acceleration of a subject, bringing him or her into close 
association with older and sexually more mature children, would 
be found associated with increased masculinity in boys and 
increased femininity in girls. Data on this point were available 
for 80 boys and 130 girls of the gifted group. The age-grade 
status of each sex ranged from one year retarded to 5 years 
accelerated, but the biserial correlation with M-F score for each 
sex was less than its probable error. Comparison of extremes 
with respect to degree of acceleration showed no significant 
difference in the case of girls. On the other hand, the 22 boys 
who were accelerated three years or more scored markedly less 
masculine than the 11 who were not at all accelerated; the 
scores were, respectively, +44 and +95. It would seem that 
in the case of boys the influence, if real, is in the opposite direction 
from the tentative assumption. Possibly the assumption was 
wrong; it may be that the boy who is excessively promoted 
becomes so isolated socially from inability to share the activities 
of his school fellows on equal terms that his mental masculiniza- 
tion is actually retarded. 

M-F SCORE AND TRAIT RATINGS 

In Chapter IV summary data were presented which resulted 
from attempts to validate M-F scores in terms of ratings of 
subjects by their teachers or acquaintances (or by themselves) on 
masculinity-femininity of personality. It was shown that the 
extremely low correlations found could be accounted for in large 



CORRELATIONS OF M-F SCORE 89 

part by the unreliability of the ratings, but that the coefficients 
would still be low even if the M-F scores were correlated 
with the composite of the ratings of several judges. If the 
validation of the M-F test hinged upon our ability to establish 
fairly high correlations with such ratings one could claim for its 
scores only the slightest validity. However, the method of 
deriving the test (by the retention only of such test items as 
actually yield sex differences in responses) automatically vali- 
dates the test as a measure of masculinity and femininity in the 
particular area (a fairly broad one) embraced by the test items 
retained. Why a measure derived from such a test should agree 
so little with subjective ratings has been a puzzling question. 

In the M-F scale we have a technique that measures mascu- 
linity-femininity in both sexes, differentiates sex groups unmis- 
takably, and rates single individuals at various points over a 
wide range. Individual scores have been found that deviate 
from the general mean as much as three or four times the S.D. 
of the distribution for the sex in question. Individual men 
have been found who rate more feminine than the average 
woman, and women who exceed the average man in mental 
masculinity. We have found that occupation, interest, age, 
education, intelligence, and homosexuality (at least in males) 
account for large deviations in group averages, and that home 
environment and marital experience are not without influence. 
It is perhaps not unreasonable to suppose that these influences 
operative in groups are discernible for individuals also, and 
furthermore that other elements in nature and nurture might be 
isolated which would account for some or part of the deviations 
of individuals within groups. One would expect that definite 
and probably rather easily distinguishable traits of character or 
personality could be described and measured which would show 
a considerable degree of correlation with the M-F scores if a 
sufficiently wide range of the latter were included. 

The low correlations found in the comparisons of scores with 
ratings on all-round masculinity-femininity suggested that the 
latter trait is perhaps too complex and vague to permit of reliable 
or valid ratings based on observation, and that higher correla- 



90 SEX AND PERSONALITY 

tions might be obtained between scores and ratings on specific 
aspects of masculinity and femininity. The terms masculine 
and feminine had perhaps been variously conceived by different 
raters; perhaps also the significance of particular behavior 
manifestations in the individuals rated had not been similarly 
interpreted. In an effort to test this hypothesis an elaborate 
correlation study was carried out which involved the rating of 
three groups of subjects on 19 different aspects of personality 
which were thought to be probably associated with the kind of 
mental masculinity and femininity measured by the test. The 19 
traits were selected partly on the basis of the clearness with 
which they could be defined, in the hope that this would make 
for agreement among judges as to the degree of their presence or 
absence. The ratings were made on a four-page blank 8j^ by 
11 in., carrying directions for rating and succinct definition of 
the traits in question (see Appendix V). 

The rating blank was first used with a small group of orphan 
children to see how it would work. The main study was directed 
to the data from sizable populations of both sexes at the eighth- 
grade and eleventh-grade levels, and of one sex at the college level. 
The eighth-grade subjects numbered approximately 75 of each 
sex, the eleventh-grade subjects approximately SO of each sex. 
The college group was composed of 92 women, members of sorori- 
ties at Stanford University. Each subject was rated on each of 
the 19 traits by either two or three judges, the school boys and 
girls by their teachers, the college women by fellow members of 
the sororities to which they belonged. For all three groups 
raters were selected who were considered to be specially qualified 
as discerners of character traits. Some, but not all of them, had 
had previous experience in rating personality traits. 

The ratings for the three groups were not, however, uniformly 
competent. Some of the teachers who rated the school boys and 
girls stated that they did not know the pupils well enough to 
feel satisfied with their ratings, even though they had had them 
in one or more classes daily for several months. This lack of 
acquaintance was indicated by the small range in the ratings 
given; the expected discrimination being evident only in the case 



CORRELATIONS OF M-F SCORE 



91 



of the more conspicuous traits of a given subject. The ratings 
of the college women were made with considerable care. A 
graduate student in psychology, Madeline Frick, assisted in 
obtaining the ratings and served as one of the raters. Miss 
Frick had given special attention to problems of personality 
and was exceptionally well equipped, both professionally and 
personally, for the task of rating the subjects herself and securing 
the intelligent cooperation of other raters. All who assisted 

TABLE 14. CORRELATION BETWEEN PERSONALITY TRAIT RATINGS AND 
M-F SCORE AT THREE AGE LEVELS 



No. 


Trait 


Women and gir.s 


Boys 


92 college 
women s 
ratings, 
standard 
scores from 
averages 
of ratings 
by 3 judges 


52 high- 
school-junior 
girls' ratings, 
averages of 
2 or 3 judges 


76 8th-grade 
girls, aver- 
ages of 2 or 
3 judges 


48 high- 
school- 
junior 
boys, aver- 
ages of 2 
or 3 judges 


78 8th-grade 
boys, aver- 
ages of 2 or 
3 judges 


1 
2 

4 

5 
6 
7 

8 
9 
10 
11 
12 
13 
14 

15 
16 

17 

18 
19 




+ 01 07 
4- 19 07 

4- 07 07 

- 05 .07 
-K10 07 
- 08 .07 

+ .03 07 
+ 15 .07 
- 08 .07 
- 04 07 
- 25 07 
4- 04 07 
4 11 07 

- 08 07 
+ .08 07 

- 03 07 

- 05 07 
- 13 07 

4- 10 07 


+ 193 09 
4- 220 09 

* 

* 
* 
* 

* 
* 
* 
* 
- 182 09 
+ 098 .095 
+ .327 .085 

* 

* 

* 

* 
* 
* 


- 111 .074 
- 079 .075 

* 

* 
* 
* 

* 
* 
* 
* 
* 
* 
* 

* 
* 

* 

* 

* 
* 


* 

* 

* 

* 
* 
* 

+ .06 .10 

* 
* 

+ .05.10 
* 
* 
4-.03 10 
4-.06l.10 

* 

* 
* 

* 


+ 24 07 
4 17 07 

* 

* 
* 

* 

* 
* 
* 
* 
* 
* 

+ 18 075 

* 

* 

* 

* 
* 
* 


Personality 
Attractiveness to 
persons of the 
opposite sex .... 
Attractiveness to 
persons of the 
same sex 


Seeks the society 
of the other sex 
Seeks the society 
of the same sex . 
Typical intellec- 
tual interests of 
own sex 


Typical intellec- 
tual interests of 
opposite sex 
Typical social in- 
terests of the 
same sex 


Typical social in- 
terests of the 
opposite sex . 
Typical manual 
interests of the 
same sex 
Typical manual in- 
terests of the 
opposite sex . 
Tomboyishness (in 
J'rls); sissiness 
n boys) 
,s crushes on 
persons of same 
sex . 


Aggressiveness. . 
Objective- minded- 
ness 
Subjective-minded- 
ness 


Effectiveness. . 
Originality ... . 



* Correlation estimated from scatter diagrams as being within the range 4-. 10 to -.10. 



92 SEX AND PERSONALITY 

were fully instructed as to rating methods, essential objectivity 
of attitude, and the necessity of avoiding "halo" effects. Three 
independent ratings were secured for the 92 women who were 
members of five sorority groups, the ratings for each group 
being made by the three of its members selected for the task on 
the basis of fitness and cooperativeness. 

The three ratings for each subject on each trait were averaged 
to secure a composite rating score. When the distributions of 
the composite scores for the different living groups were examined 
it was found that the absolute standard had evidently varied 
from house to house; accordingly the trait scores were trans- 
formed into standard scores for the several house groups. The 
resulting scores are as well equated as we could make them where 
different groups of raters provided different segments of the data. 
The correlations of M-F scores with the composite ratings thus 
treated are presented in Table 14 for the college women and 
also for the public-school boys and girls. 

A multiple correlation for 92 college women between M-F 
score and four variables (intelligence score and ratings on traits 8, 
11, and 18) is +.347 .061. Multiple correlations have also 
been computed for 78 of the 92 on M-F score with these same 
4 variables (r = +.327 .067); with these four combined 
with six others (pulse rate, and five ratios to stature, namely, 
depth of abdomen, breadth of chest, girth of ninth rib, lung 
capacity, girth of waist), r = +.443 .061; and bet ween M-F 
score and the last named six variables, r = +.358 .061. 

The correlations of ratings with M-F scores are again dis- 
appointingly low. In the case of the college women only one 
of the correlations in Table 14 is more than three times its prob- 
able error: that of .25 .07 between M-F score and "having 
typical manual interests of the same sex." Two others are 
more than twice the probable error: that with "personality" 
(+.19 .07), and that with " intellectual interests of the opposite 
sex" (+.15 .07). One other, that with "effectiveness," 
is just below this magnitude (.13 .07), in this case the 
direction of the correlation being the reverse of what most 
observers would probably have expected. For the high-school 



CORRELATIONS OF M-F SCORE 93 

girls one correlation is nearly four times its probable error, that 
of +.327 .085 for "tomboyishness." Three others for this 
group are more than twice their probable errors: "leadership" 
(+.19 .09), "force of personality" (+.22 .09), and "typical 
manual interests of the same sex" (.182 + .09). In the case 
of the eighth-grade girls and the high-school boys there is not a 
single correlation that even suggests statistical significance. For 
the eighth-grade boys, one is more than three times its probable 
error ("leadership," +.24 .07) and two are more than twice 
their probable errors ("force of personality," +.17 -07, and 
"sissiness," +.18 .075). That with "sissiness" is in the 
unexpected direction in eighth-grade boys perhaps because the 
early adolescent interest in girls of the older and hence more 
masculine-scoring boys in the group is mistakenly interpreted as 
an evidence of "sissiness." 

It will be noted that masculinity of score shows a possibly 
significant positive correlation with "leadership" in two of the 
five groups (high-school girls and eighth-grade boys) and "force 
of personality" in three of the five groups (college women, high- 
school girls, and eighth-grade boys). "Typical manual interests 
of the same sex" correlates negatively with masculinity, and 
"tomboyishness" positively, in two of the three female groups. 
All these trends are in the expected direction and, although 
each coefficient considered alone is of doubtful significance, the 
correlations fit together in a way to indicate the existence of 
small but fairly definite relationships. The small negative 
correlation between M-F score and "effectiveness," in the case 
of the college women, may not be a result of chance factors; it 
is quite possible that the kind of effectiveness the raters had in 
mind was effectiveness in activities essentially feminine in char- 
acter. The reason for the positive correlation between mascu- 
linity score and "sissiness" in the case of eighth-grade boys 
is obscure, assuming it is not a chance result. There are two or 
three circumstances, however, that taken together may account 
for it. (1) The more masculine-scoring boys are probably on 
the average somewhat older, as it has been shown that at this 
age there is an appreciable correlation between age and score in 



94 SEX AND PERSONALITY 

the case of boys. (2) The older boys at this school level may 
be growing out of the boisterous period and therefore tend on 
the average to present the picture of a quieter and less obstreper- 
ous (less masculine) personality, an effect which could easily be 
interpreted by the teacher as bordering upon "sissiness." 

The evidence with reference to the attractiveness or unattrac- 
tiveness of the more or less masculine girl or young woman is 
one of the most interesting and perhaps one of the more surprising 
contributions of this experiment. At the grammar-school and 
high-school levels the correlation between M-F score and the 
groups of traits indicating attractiveness to and interest in their 
own and the opposite sex are all zero. At the college level 
attractiveness to and interest in the opposite sex tend if any- 
thing to be associated with the more masculine M-F scores, 
while attractiveness to and interest in their own sex tend to be 
associated with the more feminine scores. In view of the fact 
that intelligence also correlates positively with the M-F score, 
these findings, even if they are accepted only to the extent that 
they do not support the opposite trends, are biologically encourag- 
ing. Similarly, it is interesting to know that in these groups 
there was no correlation, either positive or negative, between 
M-F score and proneness to form emotional attachments to 
members of the same sex, though it is unlikely that there were 
genuine cases of homosexuality in these groups. 1 

1 Of the 92 women for whom trait ratings were made at college in 1928 the 
marital status eight years later is as follows: 66 married, 23 unmarried, 2 married 
and divorced, 1 (unmarried) deceased. Comparing the 66 married with the 23 
unmarried with respect to total M-F score, intelligence score, and each of the 
19 trait ratings we find the results shown in the table on page 95. (A plus 
difference means that the women of the group who are married have the higher 
rating in the trait in question.) 

Definitely significant is the leadership difference only, possibly also that for 
objective-mindedness. A glance at the other differences shows that the married 
group tend to exceed in reasonableness, attractiveness to their own sex 
(including the raters), strength of personality, effectiveness, and having the 
typical intellectual and manual interests of the opposite sex and of their 
own. A combined personality rating for leadership, balance, attractiveness to 
the raters, and breadth of interests would give a significantly higher score to 
the married group. The three traits in which the differences have negative 
signs are, having the typical social interests of their own sex, having crushes, 



CORRELATIONS OF M-F SCORE 



95 



How shall we interpret the scantiness of positive results from 
this experiment in trait rating? Would the conclusion be 
warranted that M-F scores are almost totally uncorrelated with 
any of the various aspects of behavior which go to make up what 
is known as personality? In our opinion such a conclusion 
would be premature. We have shown in other chapters that 
the scores of this test are definitely related to occupational 
interest, mechanical interest (in the case of men), religious 
interest, culture, intelligence, age, education, aggressive inde- 
pendence, and homosexuality. We have also found that 
composite subjective rankings of both male and female groups for 
masculinity-femininity agree fairly closely with the rank orders 
of the same groups based upon mean M-F scores. 

One can only conclude either (1) that the type of masculinity 
and femininity tested by the test has few conspicuous correlates 
in everyday behavior, or (2) that these correlates cannot be accu- 
rately identified by teachers or close acquaintances. That the 
latter explanation is generally speaking correct is, we believe, 

and emotionality and excitability, but none of these differences is statistically 
significant. The differences in M-F score and in intelligence are positive but not 
statistically significant. Of the seven women who were rated above 1 in standard 



Traitf 


Diff.* 


Diff. 


S.D.diff. 


I 


+ .77 


+ 3 35 


2 


+ .38 


+ 1.67 


3 


+ .25 


+ .97 


4 


+ .40 


+ 1.76 


5 


+ .05 


+ .18 


6 


+ .07 


+ .29 


7 


+ 30 


+ 1.10 


8 


+ 42 


+ 1 71 


9 


-.04 


- 16 


10 


+ 25 


+ .88 


11 


+ .34 


+ 1 39 



Trait 


Diff.* 


Diff. 
S.D. dl f f . 


12 


+ .29 


+ 1 28 


13 


+ .05 


+ 21 


14 


- .03 


- 13 


15 


+ .20 


+ .90 


16 


+ .47 


+ 2 25 


17 


- 32 


-1 53 


18 


+ 30 


+ 1 39 


19 


+ 12 


+ 45 


M-F score . . 


+ 2 3 


+ 25 


Intelligence 
Thorndike 
score 


+ 1.7 


+ ,77 









t See Appendix V for rating sheet with names of traits. 

* The differences are in terms of ratings expressed as standard scores. 

score on "having crushes" five are married, one has been married and is divorced, 
one is single. Of the two women in the total group of 92 who rated above 2 in 
standard score on "having crushes" one is married, the other single. 



96 SEX AND PERSONALITY 

supported by our data on occupations, interests, culture, age, 
and homosexuality. The small amount of agreement between 
two or more raters who judge the masculinity-femininity of the 
same subjects is corroborative of this interpretation. In 
Chapter IV it was shown that a composite of several ratings of 
masculinity-f emininity by teachers would be necessary to afford 
even a very modest reliability. The ratings of the more specific 
aspects of masculinity-femininity just described have on the 
average higher reliabilities. These are shown in Table IS for 
eight of the traits. It will be noted that the reliabilities for 
these more specific traits run much higher than for ratings of 
all-round masculinity. Even so, we have seen that there is 
little evidence of correlation between composite ratings and 
M-F score. 

TABLE 15. RELIABILITY OF CERTAIN TRAIT RATINGS OF COLLEGE WOMEN 



1. Leadership . ... 

2. Force of personality. . 

8. Having typical intellectual interests of oppo- 



site sex. 



9. Having typical social interests of same sex 
11. Having typical manual interests of same sex 

13a. "Tomboyishness" 

14. Having crushes on persons of same sex. 
18. Effectiveness 



Effectiveness 



Reliabilities 



74 (av. of 3) 
43 (av. of 15) 

45 (av. of 15) 
45 (av. of 3) 
.39(av. of 15) 
.44(av. of 3) 
41 (av. of 3) 
59 (av. of 15) 
72 (av of 3) 



SUMMARY 

1 . No reliable correlations have been found between M-F scores 
and self-ratings on health in the case of our adult population. 

2. Most of the correlations between M-F score and physical 
measurements are within one or one and a half probable errors of 
zero. The data suggest that there may be for males a positive 
correlation between masculinity and height and also between 
femininity and length of trunk in relation to height. In the case 
of females the only significant relationship suggested is the 
positive correlation between masculinity of score and ratio of 



CORRELATIONS OF M-F SCORE 97 

girth at ninth rib to stature. No other relationships between 
M-F score and physical secondary sexual characteristics were 
suggested. 

3. It is conceivable that the apparent correlation with height 
is not the result of an intrinsically more masculine element 
expressed in relatively greater size, but rather the result of a 
correlation between masculinity of score and the kind of ascend- 
ancy and domination which is associated with superior strength 
and athletic ability; in other words, whatever relationship exists 
may be due to psychosociological rather than psychobiological 
factors. 

4. Exceptionally early puberty appears to be associated, in 
the case of boys, with excess masculinity of score in the late 
teens; in the case of girls, with excess femininity of score at 
this age. 

5. Ratings on 19 personality traits commonly believed to be 
associated with mental masculinity-femininity have yielded only 
negligible or very low correlations with M-F scores. Indications 
of small relationships are found in the case of the following rated 
traits: "Having the typical manual interests of their own sex," 
''force of personality," and " tomboy ishness." 

6. Since intelligence correlates positively with M-F score in 
women it is perhaps fortunate that "attractiveness to persons 
of the opposite sex" shows no evidence of negative correlation 
with test standing. 

7. The reliabilities of the ratings on the fairly specific per- 
sonality variables entering into this investigation are considerably 
higher than we have found for ratings of all-round masculinity, 
but except in the case of leadership are still unsatisfactorily low. 

8. Further investigation is urgently needed to throw light on 
the behavioral correlates of the kind of masculinity and femininity 
measured by this test. 



CHAPTER VI 

CORRELATION OF M-F SCORE WITH PERSONALITY AND 
ACHIEVEMENT MEASURES 

EXTROVERSION-INTROVERSION AND THE M-F SCORE 
Two measures of the habit systems which type psychology 
attempts to differentiate in an extrovert-introvert dichotomy 
have been applied and correlated with ratings on the M-F scale. 
Scores on the Laird C2 Personal Inventory 1 (higher the score, the 
more introvert) have been correlated with M-F scores (higher, 
the more masculine) for 90 men in an elementary psychology 
class (r = .25 .07), for 46 men of high (upper quartile) 
scholarship (r = .29 .09), for 46 men of low (lower quartile) 
scholarship (r = +.20 .10), and for 55 college women in a 
mixed group (r = .19 .09). The M-F scores of 329 older 
delinquent boys were correlated with the Neyman-Kohlstedt 
scores by Casselberry 2 (r = .02 .04). 

In three of the four groups for which correlations were obtained 
the coefficients indicated a low negative correlation. Two of the 
correlations for men are high enough to be fairly reliable. The 
women's correlation suggests that the same trend is present for 
them also. The third correlation for the men, although not 
definitely significant, is curiously divergent from the general 
trend indicated in the other groups, but it may be noted that both 
the M-F scores and the Laird C2 scores cover much narrower 
ranges for this group than for the others. Altogether it seems 
probable that for unselected groups of college men there is a 
small negative correlation between Laird introversion and 

1 The items included in the Laird C2 are essentially those enumerated by Max 
Freyd (Introverts and extroverts, PsychoL Rev., 1924, 31, 74-87) in so far as 
they were found valid by Edna Heidbreder (Introversion and extroversion in 
men and women, /. Abn. and Soc. Psychol., 1927, 22, 52). 

* CASSELBERRY, W. S., Analysis and prediction of delinquency, J. Juv. Res. t 
1932, 16, 1-31. 

98 



CORRELATION OF M-F SCORE WITH PERSONALITY 99 

masculinity as measured by our test, and that the same general 
trend may be present to a somewhat less extent in the women 
also. The trend is not unexpected in view of the M-F culture 
relationship: the more masculine college men are the more 
extroverted, the more feminine are the more introverted, and 
the same may be true for the women. 

A study by Oliver 1 gives corroboration to the proposition just 
stated and adds further interesting sidelights on the relation 
between extroversion-introversion and M-F score. Oliver 
selected for comparison small groups of extreme extroverts and 
introverts, the 10 per cent tail from each end of a normal distribu- 
tion of college students. The 12 extrovert and 11 introvert men 
so selected were given a number of personality tests in an effort 
to determine some of the traits associated with extreme extro- 
version and extreme introversion. On the M-F scale the 12 
extroverts rated at +108.3, the 11 introverts at +66.8. The 
difference between these two means is 5. 5 times its standard 
error. The characteristics which Oliver discovered for his 
extrovert group we may therefore fairly accept as (indirectly) the 
more masculine traits, the introvert traits as (also indirectly) the 
more feminine. An added justification for this interpretation is 
found in the relationship between M-F score and extroversion- 
introversion in high- and low-scholarship men studied by Tinsley 
(see pp. 101 /.). For them also the Laird scores and M-F scores 
were negatively correlated. The high-scholarship group with 
average Laird C2 score of 18 rated at +61.1 on the M-F scale; the 
low-scholarship group with average Laird C2 score of 14 rated at 
+92.3 M-F. Here again extroversion is associated with mascu- 
linity, introversion with femininity. The difference between 
the M-F means is 3.8 times its standard error; that between the 
Laird C2 means 3.1 its standard error. 

The other tests used by Oliver to define the traits of extroverts 
and introverts were Strong's vocational-interest test, the Kent- 
Rosanoff free association test given as a group test, the Pressey 
X-O test, Allport's A-S reaction test, the Watson test of fair- 

1 OLIVER, R. A. C., The traits of extroverts and introverts, /. Soc. Psychol., 
1930, 1, 343-665. 



100 SEX AND PERSONALITY 

mindedness, and the George Washington test of social intelli- 
gence. Comparison of Oliver's (masculine) extroverts with his 
(less masculine) introverts on these various tests suggests at 
least what may be expected from a direct comparison between 
the measures used and the masculinity-femininity score. The 
IS 1 extrovert men showed less vocational aptitude, as measured 
by Strong's test, than the 13 l introvert men for the occupations 
of their choice. These were mechanical engineering, personnel 
psychology, law, advertising, and architecture. The introverts 
chose journalism, law, authorship, medicine, advertising, execu- 
tive administration, and insurance. It is of interest to note that 
the extrovert who chose to be a mechanical engineer was the only 
one in the extrovert group who received a rating of A in the 
Strong test for the occupation of his choice. None of the group 
rated B+. Among the introverts the choices of journalism, law, 
and authorship received Strong ratings of A; medicine, author- 
ship, and law, B+. Furthermore, the extroverts were not only 
less well suited to the professions they had chosen, but it was 
also found that fewer of them than of the introverts had made a 
decision in the matter. This agrees with our findings reported 
in Chapter VIII, that the more masculine men have less intense 
positive interests than the less masculine. The contrast between 
mechanical interest and culture is further brought out by the 
ratings of all the members of both groups on the two occupations 
"author" and "engineer." The former occupation proves by 
this comparison to be, as we might expect, quite definitely 
introvert and less masculine, the latter extrovert and more 
masculine. 

In the Kent-Rosanoff free association test both of Oliver's 
groups gave a larger percentage of unique responses than the 
norm, as was to be expected in view of their more than average 
education. But the introverts (less masculine) gave significantly 
more of the unique responses than the extroverts, whatever that 
may mean. The introverts showed on the Pressey X-0 tests 

1 Three subjects (extreme extroverts) were added by Oliver to his original 
group of 12 extroverts, and two extreme introverts to his original group of 11 
introverts. 



CORRELATION OF M-F SCORE WITH PERSONALITY 101 

higher affectivity, greater idiosyncrasy, and greater self-con- 
sciousness; the last two differences are statistically significant. 
These results agree with expectation regarding introverts and 
extroverts, and also with the culture interpretation of femininity 
in men and its contrast in the indifferent, hard-boiled, over- 
masculine type. Allport's A-S reaction test showed the extro- 
verts to be more ascendant, the introverts more submissive than 
the norm. The difference is not entirely reliable, but the direc- 
tion in the contrast is worthy of note. Since the extroverts 
are more masculine, it is not surprising that they should also 
be more ascendant. The Watson test of fair-mindedness 
showed no difference between the two groups in general level of 
prejudice, and none in any particular aspect of religious opinion. 
There appeared, however, to be a definite tendency for introverts 
to hold more liberal views on economic issues and to favor 
stricter moral standards. On the social-intelligence test there 
was no difference found between the two groups. 

Additional data with reference to the relation of M-F score to 
Conklin scores on introversion-extroversion come from Tinsley's 1 
study of the traits of high- and low-scholarship men. Her high- 
scholarship men with low M-F mean (+61.1) rated introvert; 
her low-scholarship men with high M-F mean (+92.3) rated 
extrovert. The difference between the two groups on Laird C2 
score was 3.1 times its standard error. The correlation between 
Conklin introversion score and M-F rating was .52 .07 for 
the high-scholarship men and .24 .09 for the low-scholarship 
men, which agrees in direction with the Laird C2 results, but 
brings out also a difference that should be further investigated; 
namely, the higher correlation of M-F with Conklin than with 
Laird. 

Scores on the tests secured by Tinsley have been correlated by 
us with the other measures in her battery (Table 16) and we 
quote in addition her own correlations and averages (Table 17). 
These results, especially the contrasting traits of the high- 

1 TINSLEY, RUTH, Personality traits of high and low scholarship men with 
equal aptitude ratings, M.A. thesis directed by P. R. Farnsworth, Stanford 
University, 1930, unpublished. 



102 



SEX AND PERSONALITY 



scholarship (more feminine) group, when compared with the 
low-scholarship (more masculine) group, corroborate Oliver's 
results from another angle. The more feminine, high-scholarship 
group differs from the more masculine, low-scholarship group 
in being (1) more submissive in terms of Allport 's test (critical 
ratio 2.7), (2) less prejudiced in terms of Watson's test (critical 
ratio 2.1), (3) more self-conscious, having higher affectivity 
scores in terms of the Pressey X-O test (critical ratio 1.8), (4) 
more unstable emotionally as measured by the Laird B2 personal 
inventory (critical ratio 1.3), and (5) as we have already seen, 

TABLE 16. CORRELATIONS BETWEEN PERSONALITY SCORES AND M-F 
SCORES IN HIGH- AND LOW-SCHOLARSHIP GROUPS 





46 college men of 
high scholarship 


46 college men of 
low scholarship 


Laird B2... 


- 31 09 


00 4- 10 


Laird C2... 


- 29 09 


4- 20 10 


Conklin 


52 .07 


24 09 


Pressey X-O 
Allport A-S 


+ .03 .10 
-h 36 .09 


-.44 .08 
+ 08 10 


Watson fair-mindedness . 


+ .18 .10 


-.17 10 



TABLE 17. CORRELATIONS BETWEEN PERSONALITY SCORES AND SCHOLARSHIP 
SCORES (GRADE-POINT AVERAGE) 



Laird B2 


+ 11 .06 


Good scholarship (indirectly femininity) 


Laird C2 . 


+ .27 .06 


and instability associated to this extent. 
Good scholarship (indirectly femininity) 


Conklin 
Pressey X-O 


+ .23 .06 
+ .13 .06 


and introversion associated to this 
extent. 
Good scholarship (indirectly femininity) 
and introversion associated to this 
extent. 
Good scholarship (indirectly femininity) 


Allport A-S . 


- 25 .06 


associated with affectivity and self- 
consciousness to this extent. 
Good scholarship (indirectly femininity) 


Watson fair-mindedness . 


-.18 .06 


and ascendancy associated to this 
extent, i.e., negatively. 
Good scholarship (indirectly femininity) 
and prejudice associated to this extent, 
i.e., negatively. 



CORRELATION OF M-F SCORE WITH PERSONALITY 103 

more introverted (critical ratio Laird, 3.1; Conklin, 2.7). 
Tinsley's correlations between scholarship and the various 
measures are a further indirect evidence of the correlation 
between these measures and the M-F score. The correlations 
between M-F score and each of the other measures suggest the 
desirability of further investigation using unselected groups with 
wide score ranges. 

In summary it may be said that there is considerable evidence 
that more masculine college men tend to be extroverted, more 
feminine college men to be introverted. This is true with 
respect to the elements of extroversion-introversion measured 
in the Laird C2 test and even more clearly the case with respect 
to the more restricted type of extroversion-introversion measured 
by the Conklin test. 1 Furthermore, the (masculine) extrovert 
men are found to be less sure of the vocations they should follow 
than the (feminine) introvert men. They are less well fitted for 
those they choose, unless these be in the field of engineering. 
The (introvert) feminine group is more affective and self- 
conscious in response on the Pressey test and shows greater 
idiosyncrasy. Its members are more submissive, hold more 
liberal economic views, and favor stricter moral standards. The 
two groups do not differ in fair-mindedness nor in social 
intelligence. That the traits in question tend to persist as 
characteristic of the more masculine and the more feminine 
groups respectively is shown by a contrast between high- (femi- 
nine) and low- (masculine) scholarship groups. This comparison 
adds to the traits already mentioned a difference in prejudice: the 
data indicating that prejudice is more characteristic of men 
rating high than of men rating low in masculinity. 

College women seem to show a slight tendency for introversion 
to be associated with femininity, extroversion with masculinity, 
but further evidence is needed to confirm or disprove the slight 
trend suggested. In teen-age delinquent boys there is no 
relation between M-F score and introversion-extroversion as 
measured by the Neyman-Kohlstedt test. 

1 The Bernreuter test of introversion has not been correlated with M-F scores, 
but it is known to correlate very highly with Laird C2. 



104 SEX AND PERSONALITY 

CORRELATION or M-F SCORE WITH CHARACTER MEASURES 

Information concerning the relation between character traits 
and masculinity-femininity has been contributed by Casselberry, 
who compared M-F scores of 329 older delinquent boys with 
their standing in the series of character tests developed by Cady 
and Raubenheimer and used in the Stanford study of gifted 
children. The battery included two overstatement tests, tests of 
questionable reading preferences and character preferences, a 
trustworthiness test (the tracing of circles and squares with 
opportunity for cheating), and Cady's adaptation of the Wood- 
worth emotional-stability test. 1 The weighted character index 
developed from these tests by Terman and Goodenough was also 
derived for the delinquent boys. Scores on the separate parts 
of the battery and the final indices were correlated with the 
M-F scores. 

The first overstatement test, which consists in marking book 
titles, some of them fictitious, gave a correlation of +.19 .04 
between M-F score and amount of overstatement regarding 
books claimed to have been read. Overstatement B, the second 
test in the series, gives an opportunity for overstatement with 
reference to general information. The correlation coefficient 
for this test and the M-F is +.10 .04, showing a less clear 
relation than was found for overstatement A. Questionable 
reading preferences and questionable character preferences, both 
of which rank very low in their contribution to the general 
character index, show no correlation with M-F score; the coeffi- 
cients are +.00 .04 and -.01 .04. The activity interest 
test, also based upon questionable preferences, shows the small 
but reliable correlation with masculinity of +.18 .04. This 
test is heavily weighted in the character index. The trust- 
worthiness test adapted by Cady and used in accordance with 
the technique developed for the gifted-children study shows a 
correlation of +.06 + .04 with M-F scores for the delinquent 
group. The Woodworth-Cady emotional-stability test gives a 

1 The tests and testing procedure are described in Terman, et al., Genetic 
studies of genius, vol. I, Chapter XVII, 1925. 



CORRELATION OF M-F SCORE WITH PERSONALITY 105 

correlation of +.04 .04 with M-F. The weighted character 
index which combines scores from all seven tests in the battery 
correlates +.06 .04 with the M-F test, showing no appreciable 
relationship between masculinity-femininity and the character 
index as a whole. Such correlation as is indicated appears to be 
limited to (1) overstatement with reference to reading matter, 
and (2) choice of questionable or undesirable activities, both of 
which are natural opposites to the more introvert and cultural 
pursuits of the boys who tend toward femininity on the M-F 
scale. 

MECHANICAL ABILITY AND MASCULINITY 

There was no more striking result in the occupational and 
interest comparisons (Chapters VIII and IX) than the influence 
of mechanical pursuits on the M-F score at every educational 
level. It is no wonder then that a positive correlation is found 
between M-F ratings and scores on a test of mechanical ability. 
It is perhaps also to be anticipated that this positive relationship 
will appear for males only, as nothing equivalent to the high 
means of the engineers and architects and the mechanical 
occupations appeared for the women. The first correla- 
tion between mechanical ability (Stenquist I) and M-F 
test scores was noted in the small group of 15 orphan boys 
(rho = +.55 .13). For 15 orphan girls the correlation 
coefficient was + .03 + .20. The Stenquist test was then given 
to 64 boys and 64 girls in the junior year of high school, the 
resulting correlations with M-F scores being as follows: for the 
boys, +.237 .079; for the girls, +.055 .083. Correlations 
between Stenquist scores and the scores on the separate parts 
of the M-F test, Exercises 3, 4, and 5, showed that Exercise 5 
(interests; r = +.20 .08) was contributing the major part 
of the correlation with the M-F score. The coefficients for 
the boys between scores on the other exercises and the Stenquist 
scores were all less than their probable errors, as were also those 
for the girls on each of Exercises 3, 4, and 5. 

A further group consisting of 329 delinquent boys in the teen 
age was given the Stenquist test by Casselberry, its scores 



106 SEX AND PERSONALITY 

correlating +.306 + .038 with M-F scores. The McQuarrie 
mechanical-ability test given by Casselberry to the same group 
yielded a correlation of +.125 .038 with M-F score. 

From these findings it may be stated that mechanical ability 
measured by test is a masculine trait for boys and young men 
as it is for older men when measured by occupation. It has no 
correlation whatever with mental masculinity-femininity in 
girls, and this agrees with the apparent absence of any such 
relationship in the occupational findings for adult women 
(Chapter VIII). 

SCHOLARSHIP AND M-F SCORE 

We have investigated the relation of scholarship to mascu- 
linity-femininity only in the case of college students. The 
factors that contribute to college marks are so numerous that one 
could hardly expect a very high correlation between scholarship 
and M-F score. Intelligence, which is more closely related to 
scholarship than any other single factor, is positively correlated 
with mental masculinity in the case of college women but not to 
any appreciable extent in the case of men. However, several 
investigations have shown that college men who rate as intro- 
verts on the Laird or Bernreuter tests display a small but reliable 
superiority in scholarship over men who rate as extroverts, and 

TABLE 18. SCHOLARSHIP AND M-F SCORES 

A. Average M-F scores of college men above and below (1) the college scholarship 
mean, (2) the college intelligence mean. 

Mean 
M-F score 

1 . 69 men, intelligence below average, scholarship below average . +77 . 

2. 40men, intelligence above average, scholarship below average. +74 2 

3. 44 men, intelligence above average, scholarship above average . +71.7 

4. 39 men, intelligence below average, scholarship above average . +77.5 
Combinations of the above: 

a. Scholarship comparison 

83 men above average in scholarship +74.4 

109 men below average in scholarship +75 . 8 

b. Intelligence comparison 

84 men above average in intelligence +72 . 8 

108 men below average in intelligence +77, 1 



CORRELATION OF M-F SCORE WITH PERSONALITY 107 

TABLE 18. SCHOLARSHIP AND M-F SCORES. (Continued) 

B. Average M-F scores of college men (1) in the upper and lower scholarship 

quartiles and (2) above and below the college intelligence mean. 

Mean 
M-F score S.D. S.D.M 

1. 43 men, high intelligence, low scholarship +89 9 43 .2 6.6 

2. 53 men, low intelligence, low scholarship .... +88 1 42 58 

3. 62 men, high intelligence, high scholarship . 4-74.4 41.8 53 

4. 27 men, low intelligence, high scholarship. . +52 9 43.8 84 
Combinations of the above: 

a. Scholarship comparison 

89 men in the upper scholarship quartile. . 4-67 8 
96 men in the lower scholarship quartile . 4-88 6 

b. Intelligence comparison 

105 men above average in intelligence . . . 4-80.6 
80 men below average in intelligence . . + 76 1 

C. Average M-F scores of college men who have been in college 6 or more terms 

(2 college years or more) and who are (1) in the upper and lower scholar- 
ship quartiles and (2) above and below the college intelligence mean. 

Mean 
M-F score S.D. S.D. M 

1. 28 men, intelligence above college average, 

scholarship in lowest quartile -f-100 5 43 8 83 

2. 18 men, intelligence below college average, 

scholarship in lowest quartile 4- 81 1 36 6 86 

3. 27 men, intelligence above college average, 

scholarship in upper quartile 4" 69 . 2 42 . 4 82 

4. 19 men, intelligence below college average, 

scholarship in upper quartile 4- 52.9 492 11 3 

Combinations of the above: 

a. Scholarship comparison 

46 men in the upper scholarship quartile . . 4- 61 . 1 46 68 
46 men in the lower scholarship quartile. ... 4-92.3 419 62 

b. Intelligence comparison 

55 men above average in intelligence. . .4-87.4 
37 men below average in intelligence 4-66.6 

since introversion, as we have seen, is for men negatively 
correlated with mental masculinity, it would not be surprising to 
find a small negative correlation between scholarship and 
masculinity. The culture factor which in men seems to be 
associated with low M-F score would support the same expecta- 
tion. In the case of women the expectation is reversed, since 
with them both superior intelligence and superior culture are 
associated with mental masculinity. Table 18 gives mean M-F 



108 SEX AND PERSONALITY 

scores of college men in their relation to both scholarship (as 
measured by grade-point ratio) and intelligence (as measured by 
Thorndike's college aptitude test). 

It will be noted that Table 18 contains three sections, A, B, 
and C. Section A compares the M-F mean of all who are above 
average in scholarship with the mean of all who are below, 
separately for groups of high and low intelligence. Section B 
makes the same comparisons between highest and lowest quartiles 
with respect to scholarship. Section C is like B except that in 
this case the data are confined to the men whose scholarship 
quartile classification could be based upon at least two full 
years of college work, whereas in A and B the scholarship record 
was for varying periods from a third of a year to three and 
two-thirds years. 

Section A of the table shows practically identical M-F means 
for men above and men below average scholarship (74.4 and 
75.8) where intelligence is disregarded. When the same subjects 
are divided into two groups according to intelligence, the lower 
intelligence group rates slightly more masculine; 77.1 as com- 
pared to 72.8. The difference is not reliable but is in the expected 
direction. When the comparison is between upper and lower 
scholarship quartiles (Section B of the table) a relationship of 
considerable magnitude is found between scholarship and M-F 
score, and this relationship obtains in both intelligence groups. 
When 'the scholarship quartile classification is based upon 
grade-point average for two years or more (Section C of table) 
the relationship is still further enhanced. We find here a mean 
M-F score of 92.3 for the lower scholarship quartile as compared 
with 61.1 for the upper scholarship quartile. The corresponding 
means for those of more than average intelligence are 100.5 and 
69.2; for those below average intelligence, 81.1 and 52.9. It 
appears, therefore, that whatever the level of intelligence there 
is a considerable negative correlation between scholarship and 
mental masculinity in the case of men. The matter deserves to 
be investigated with larger groups, for if the trends here shown 
should be confirmed the M-F test might prove a useful aid in the 
prediction of college success. 



CORRELATION OF M-F SCORE WITH PERSONALITY 109 

The interests expressed by Tinsley's high- and low-scholarship 
groups (Table 19) throw light on the grade-getting success of the 
one and the lack of it in the other. The averages for both 
groups having each of the 12 interests, positive and negative, 
follow the usual trends. In general the devotees of any given 
interest have a mean M-F score above or below the mean of the 
group as a whole according as the interest in question is mascu- 
line or feminine (see Chapter DC). For example, the men who 

+ 1 00-}*" H '9 n Intelligence, Lowest Quartile Scholarship 

Unselected Intelligence, Lowest Quartile Scholarship 
+90 

High Intelligence, Unselected Scholarship 



+80 



t60 



Low lntelli9ence.LowestQuartile Scholarship 



**- High Intelligence, Highest Quartile Scholarship 



Low Intelligence, Unselected Scholarship 
Unselected Intelligence, Highest Quartile Scholarship 



<- Low Intelligence, Highest Quartile Scholarship 
+ 50 1 
FIG. 2. M-F scores of men in relation to intelligence and scholarship. 

care much for mechanics in the high-scholarship group are as 
masculine as the mechanically interested in the low-scholarship 
group. The single individual in the low-scholarship group who 
is interested in religion has a very feminine score indeed. Com- 
parison of the percentages of each group that shows positive 
interests in the various pursuits gives some indication of the 
characteristic contrast between the individuals of high and low 
scholarship. The high-scholarship group is more interested 
than the low-scholarship group in (1) politics, (2) travel, (3) 
religion, (4) literature. The differences are not large, but taken 
together they quite clearly indicate a cultural trend in the high- 
scholarship men. The negative interests of these men are social 
life, mechanics, music, pets, and in these a larger percentage of 
them than of the low-scholarship men have little or no interest. 



110 



SEX AND PERSONALITY 



It may seem surprising that music is included here, but we are 
inclined to think that musical interest among these western 
college students may not be really cultural. It is perhaps 
connected rather with sports and social life than with cultural life. 

TABLE 19. MEAN M-F SCORES OF HIGH- AND LOW-SCHOLARSHIP MEN HAVING 
(1) MUCH, OR (2) LITTLE OR No INTEREST IN CERTAIN PURSUITS 



Much interest 


Little interest 




High- 


Low- 


High- 


Low- 


Interest 


scholarship 


scholarship 


scholarship 


scholarship 




group 


group 


group 


group 


M-F means of groups 


+61.1 


+92.3 


+61.1 


+92.3 






M-F 




M-F 




M-F 




M-F 




N 




N 




A* 




N 








mean 




mean 




mean 




mean 


Travel 


U 


59 9 


31 


93 5 










Outdoor sports 


22 


69 5 


21 


95.4 


1 


65.5 


2 


89 5 


Religion 


3 


55 5 


1 


24 5 


18 


69 9 


29 


98 2 


Mechanics 


6 


118 8 


15 


99 8 


12 


39 6 


2 


72.0 


Social life 


8 


70.5 


12 


96 1 


4 


63.0 






Literature 


15 


54 8 


14 


80 2 


3 


142 1 


4 


114 5 


Music. ... 


15 


40 8 


18 


88 9 


4 


110.5 


2 


84 5 


Art 


? 


5 5 


5 


84 5 


10 


96 5 


11 


123 5 


Science . ... 


15 


84.8 


18 


97 8 


1 


55.5 


5 


86 5 


Politics . 


12 


73 8 


7 


73.0 


10 


72 5 


14 


83 7 


Domestic art. 










23 


67.6 


22 


95 4 


Pets 


7 


64 


9 


76 7 


11 


57 3 


9 


103 3 



The low-scholarship group (Table 20) has a slightly stronger 
inclination to both positive and negative interests. The differ- 
ence is small, 2 per cent for each interest in both groups of 
interests. The positive interests of the low-scholarship group 
show a strong predilection for mechanics, social life, art, music, 
science, and pets, in that order. Their negative interests show a 
striking dissimilarity as compared with those of the high- 
scholarship group with respect to attitude toward religion. The 
other pursuits for which they have expressed less interest than 
the high-scholarship group are science and politics. Again it 
may be seen that the intellectual pursuits as a whole, excepting 



CORRELATION OF M-F SCORE WITH PERSONALITY 111 



art and music, are more attractive to the high- than to the low- 
scholarship group. The outstanding characteristic of the 
high-scholarship group is its significantly smaller number of 
individuals who have very great interest in mechanics. The 

TABLE 20. POSITIVE AND NEGATIVE INTERESTS OF HIGH AND Low SCHOLARSHIP 
GROUPS or COLLEGE MEN 





High- 
scholarship 


Low- 
scholarship 




High- 
scholarship 


Low- 
scholarship 






group, 
per cent 
with 


group, 
per cent 
with 


Diff. 


group, 
per cent 
with 


group, 
per cent 
with 


Diff. 




positive 
interests 


positive 
interests 




negative 
interests 


negative 
interests 




Sports 


48 


46 


? 


2 


4 


- ? 


Social life... 


17 


26 


- 9 


9 





9 


Mechanics. . 


13 


33 


-20 


26 


17 


9 


Science ... . 


33 


39 


- 6 


2 


11 


- 9 


Literature 
Travel 


33 
74 


30 
67 


3 
7 


7 



9 



- 2 



Pets 


15 


20 


- 5 


24 


20 


4 


Politics. .. 


26 


15 


11 


22 


30 


- 8 


Art 


4 


11 


- 7 


22 


24 


- ? 


Music 


33 


39 


- 6 


9 


4 


5 


Religion 
Domestic arts 


7 


2 


5 


39 
50 


63 
48 


-24 
2 


Average per cent 


25.2 


27.3 




17.7 


19.1 





striking characteristic of the low-scholarship group is the signifi- 
cantly larger percentage of individuals who have little or no 
interest in religion. On the basis of these expressions of interest 
alone it could have been predicted that the high-scholarship 
group would register less masculine on the M-F scale than the 
low-scholarship group. 

Among college women there is evidence of about the same 
degree of correlation between scholarship and the M-F score as 
among the men. At each of two intelligence levels scholarship 
above average is associated with a more feminine score as 
follows: 



112 SEX AND PERSONALITY 

Mean M-F 
score 

1. 22 women, intelligence above average, scholar- 

ship below average 13.5 

2. 93 women, intelligence above average, scholar- 

ship above average 26.9 

3. 24 women, intelligence below average, scholar- 

ship below average 41 . 3 

4. 36 women, intelligence below average, scholar- 

ship above average 49 . 6 

Combinations of these scores show, however, that the scholar- 
ship element is small in comparison with the contribution of 
intelligence : 

Mean M-F 
score 

a. Scholarship comparison 

129 women, above average in scholarship. . . . 33.1 
46 women, below average in scholarship 27.9 

Difference ... + 5.2 

b. Intelligence comparison 

115 women, above average in intelligence 24. 3 

60 women, below average in intelligence 46.2 



Difference.... -21 9 

Low scholarship for the women, as for the men, is associated 
with more masculine M-F scores, and this persists in the more 
intelligent (more masculine) and in the less intelligent (more 
feminine) college women's groups. This seems especially 
interesting and important in view of the fact that intelligence 
operates in the opposite direction for the women (i.e., makes for 
masculinity), while for men intelligence shows no consistent 
relationship to M-F score. 

RELATION OF M-F SCORES TO ATHLETIC ABILITY 

The scores of a group of 43 male college athletes average 
+92.5, which is reliably more masculine than the mean (+67.4) 
for college men in general. They are more masculine than 
groups of either high or low scholarship unselected as to athletic 



CORRELATION OF M-F SCORE WITH PERSONALITY 113 



ability. They are also more masculine than either high- or 
low-intelligence groups unselected as to athletic ability. The 
one-third of them who have intelligence scores above average 
score +95.0; the two-thirds whose intelligence scores are below 
average score +98.5. The half whose scholarship is above 
average rates +91.0, or .25 s.s. 1 higher than the low-scholarship 
-10 



-20 



-30 



-40 



-50 



-60 



Superior College Women 

Intelligence above average of their group, scholarship below 



Superior College Women 

Intelligence above average of their group, scholarship above 



Superior College Women 

Intelligence below average of their group, scholarship below 

Superior College Women 

Intelligence below average of their group, scholarship above 



*~ Unselected College Sophomore Women 



70-U- College Educated Women Adults,aged 20 to 30 
PIG. 3. M-F scores of women in relation to intelligence and scholarship. 

half, which rates at +104.0. This is in line with expected 
cultural and personality differences found to be correlated 
with the M-F score. 

Two sets of data on women athletes are available: 
1. A group of S3 college athletes from two higher institutions 
in the West, women of exceptional skill in games and sports, 
yields a mean M-F score of 21.7. This is almost one standard 
deviation above the generality of college women. 2 

1 See Appendix II for table of standard scores for adult population. 

"Scores were also secured for 25 physical education majors in a middle- 
western university. These show an M-F average at just the same level as the 
130 college sophomores upon whom the college norm is based. But it may be 
noted that these physical education majors, many of them, are interested in the 
artistic rather than the athletic side of the profession. Hence, we incline to 
ascribe the difference between these scores and the athletic women's scores, not 



114 SEX AND PERSONALITY 

2. At Stanford University, with the assistance of Miss Helen 
M. Bunting, Associate Professor of Physical Education, and 
three women athletes of recognized standing and wide acquaint- 
ance, ratings of athletic ability were made on 100 women students. 
The individuals rated included about 70 women who had played 
on teams and about 30 women chosen at random from the list 
of others for whom M-F scores were available. Ratings were 
made on a seven-point scale the items of which are as follows: 

1. (a) Exceptional all-around athlete, very unusual skill and ability, 

outstanding in several sports, (b) Eminent athlete distinguished 
as an exceptional record holder or champion in one sport. 

2. Very skillful athlete but not most eminent. Member of first team or 

teams, consistent point winner. 

3. Fair athlete above average of college women. Dependable mem- 

ber of second team. Somewhat less able than average member 
of first team. May be substitute on first team. 

4. Average athletic ability among college women. Exercises whether 

requirements demand it or not. Likes out-of-door activity and 
regularly indulges in it. 

5. Plays games in gymnasium but does not go out voluntarily for sports. 

Chooses the less active physical education requirements. 

6. Not athletic, utmost exercise is walking. Avoids games. 

7. Avoids physical exertion entirely. 

x Means you do not know the person. 

The resulting distribution was naturally much skewed in the 
direction of the superior athletes. The small group of non- 
athletic individuals served as a satisfactory check in the rating, 
and the ranks obtained by these women indicated by comparison 
that the athletic ability of the others was probably not overrated. 
On the basis of the combined ratings of the four judges 55 women 
were found who rated 2.0 or better in athletic ability. The 
average of the M-F scores of these exceptional athletes was 
25.8, which has a standard score index of +1.40 and is +.85 s.s. 
above the college sophomore norm for women. 

In order to isolate the factor of athletic achievement or interest 
from other masculine trends in women the athletes in each of four 

to a geographical difference, but rather to the probability that a good many of 
the physical education majors included here are not athletes, but dancing 
experts. 



CORRELATION OF M-F SCORE WITH PERSONALITY 115 



groups were compared with the group as a whole as follows: (1) 
those above the college average in intelligence and in college 
grades, (2) those above the college average in intelligence and 
below it in college grades, (3) those below the college average 
in intelligence and above it in college grades, (4) those below the 
college average both in intelligence and in college grades. 
"Above average intelligence" means a score of 80 or above on 
the Thorndike college aptitude test, "below average intelligence" 
means rating at 79 or below. "Above average in college grades " 
means having a grade point rating of 1.5 or better; "below 
average in college grades," 1.49 or lower. 

TABLE 21. M-F SCORES OF ATHLETIC AND NONATHLETIC COLLEGE WOMEN 





N 


Mean 
M-F score 


S.D.M 


1. AI college athletes, total group 


44 


-21 1 


7 9 


a. with aptitude score of 80 or higher 


30 


-11.5 


8.6 


b. with aptitude score below 80 


14 


-38.1 


16 4 


2. Nonathletic college women, total group. . 


122 


-37 1 


3 6 


a. with aptitude score of 80 or higher . . 


77 


-31 


4 2 


b with aptitude score below 80 


45 


-47 6 


6 5 











In each of the four subgroups the superior athletes of that 
group score more masculine than the group as a whole except in 
the case of the women of low intelligence and high scholarship. 
Here the 14 superior athletes included score .2 of a standard score 
more feminine than the group as a whole. Although the groups 
compared are not large, it seems possible that this difference 
is not due to chance. We suggest that these women of lower 
intelligence who have made good in both scholarship and athletics 
in such an exceptional way possess a trait of effective conformity 
which is probably mentally feminine. We have found in 
another experiment that the mean M-F score of college women 
who rated below average in "effectiveness" was 29.3, as 
compared with 38.8 for those above average in "effectiveness." 
The difference between the two averages is just more than its 
standard error. 



116 



SEX AND PERSONALITY 



It seems not improbable that effectiveness in women represents 
the opposite of independence, that it involves an element of 
conformity and cooperation which is feminine in its influence. 
If this is so it is also not purely a matter of chance that the 
athletes below the college average in intelligence and above it in 
scholarship represent the most feminine extreme, while the 



+ 10 



-10 



-20 



-30 



-40 



-50. 



Mole Clergymen f 

Al Athletic College Women I 

Male Artists l 



Total Male Population, Aged 70-80 



f 

College Women-' 

- AI Athletic College W< 



/High Intelligence 
flLow Scholarship 



/omen 



Al Athletic College Women 



Low Intelligence 
Low Scholarship 
High Intelligence 
High Scholarship 



,. ,, , /High Intelligence 

College Women | Hf | h Scholarship 



- ,, XA/ /Low Intelligence 

College Women | Low Scholarship 



College Women 



-60 s- Al Athletic College Women 
FIG. 4. M-F scores of athletic and other college- women's groups. 

athletes above the college average in intelligence and below it in 
scholarship represent the most masculine extreme. We believe 
that the same indifference and independence of this last-named 
gifted galaxy makes them poor scholars and at the same time 
the most masculine group we have discovered among more than 
2,000 women tested. These women doubtless lack the feminine 
"effectiveness" which is associated with "having the manual 
interests of their own sex," and which is associated with making 



CORRELATION OF M-F SCORE WITH PERSONALITY 117 

good college marks. They are able, and they are independent. 
In a sense, this small group of seven unusual individuals epito- 
mizes the characteristics of the mentally masculine college 
woman: (1) superior intelligence, (2) marked indifference to 
traditional female college aims, (3) athletic prowess. This is the 
only female group whose mean M-F score overlaps the normal 
male range of means. These women, who are 2.25 s.s. removed 
from the mean of their own sex, are only .45 s.s. below the male 
mean. It would probably be a surprise to all in the following 
groups to know that these unusual women rate at about the same 
point on the M-F scale as male clergymen, and the generality 
of males aged 60 to 70, or that they rate more masculine than 
male artists or the generality of males above age 70. 

RELATION OF MASCULINITY-FEMININITY TO MARITAL 
COMPATIBILITY 

It would not be unreasonable to suppose that a relationship 
might exist between marital happiness and the mental mascu- 
linity or femininity of spouses. Inquiry discloses that various 
opinions are prevalent in regard to the types of M-F matings that 
would probably be most favorable to happiness. A majority of 
those to whom the question has been put are inclined to believe 
that the best combination is superior masculinity of husband 
and superior femininity of wife, though a few have argued 
strongly for the mating of "likes," some even in favor of the 
feminine-husband, masculine-wife combination. 

It has not been possible to investigate this question with the 
present M-F test, but an investigation was carried out by 
Terman and Buttenwieser 1 with the use of a substitute M-F 
test. The reader will recall that certain items in the M-F test 
were taken from the Strong test of occupational interests, 
particularly items in Exercise 5 relating to interests. It hap- 
pened that Terman and Buttenwieser had administered the 
Strong test to 341 married couples and to 100 divorced couples, 

1 TERMAN, LEWIS M., and PAUL BUTTENWIESER, Personality factors in marital 
compatibility, /. Soc. Psychol., 1935, 6, 143-171, 267-289. 



118 SEX AND PERSONALITY 

and that for each of the married couples a marital-happiness 
score was available based upon confidential information fur- 
nished by both spouses. It seemed desirable, therefore, to 
derive a masculinity-femininity score from the Strong test for use 
with these groups. 

At our suggestion an M-F scoring key was worked out for the 
Strong test by Dr. Harold Carter. The key was based on the 
responses of 114 males and 114 females to the Strong test, a 
majority of the subjects being of high-school age. Of the 
420 items in the test, 156 yielded sex differences of probably 
significant magnitude, and weights were assigned to these items 
in proportion to the amount and reliability of difference found. 
The score derived by the use of this key will be designated the 
occupational M-F score. 1 It is not identical with the Terman- 
Miles M-F score but correlated with it to the extent of .43 for 
41 males and .62 for 62 females. Any relationship that may 
be found to hold between the occupational M-F score and 
marital happiness will at least suggest the direction of relation- 
ship that would be found for the Terman-Miles M-F score. 

First the husband-wife correlation in occupational M-F score 
was found for 126 of the married couples having high happiness 
ratings, and also for 215 married couples with low happiness 
ratings. These were as follows: for the more happy group, 
.029 .060; for the less happy group, .085 .046. In other 
words, happy spouses were no more and no less alike in mascu- 
linity-femininity than unhappy spouses; marital selection based 
upon this personality trait had not occurred to any appreciable 
degree in either group. 

Next, the husband's happiness score was correlated with his 
score on the occupational M-F test (.085 .031), and the wife's 
happiness score with hers (.047 .031). Again no significant 
relationship with marital happiness appears. The result is the 
same when the combined happiness score of husband and wife 
is correlated with the husband-wife difference in Strong M-F 
score (r = .108 .03). 

1 For a later version of the Strong M-F test see Edward K. Strong, Jr., Interests 
of men and women, /. Soc. Psychol., 1936, 7, 49-67. 



CORRELATION OF M-F SCORE WITH PERSONALITY 119 



Finally, a comparison was made of the mean occupational 
M-F scores, separately by sex, for the following groups: 100 most 
happily married couples, 100 most unhappily married couples, 
and 100 divorced couples. The results, in terms of mean 
standard scores on the occupational M-F test, were as follows: 





Husbands 


Wives 


Happy 


Unhappy 


Divorced 


Happy 


Unhappy 


Divorced 


Mean occupational 
M-F standard score. 

<TM 


4.67 
20 


5.14 
.19 


4.29 
18 


4.71 
.24 


4.56 
.21 


5.45 
20 





The critical ratios of the differences between M-F scores of 
happy, unhappy, and divorced are as follows, a plus sign indicat- 
ing a higher mean standard score for the first member of the pair 
concerned, a minus sign a lower score. 



Husbands 


Wives 


Happy- 
unhappy 


Happy- 
divorced 


Unhappy- 
divorced 


Happy- 
unhappy 


Happy- 
divorced 


Unhappy- 
divorced 


-1 67 


-hi 41 


+3 24 


+0 47 


-2.39 


-3.08 



Here the outstanding difference among men is the markedly 
masculine rating of the unhappy group as compared with the 
divorced group, the happy group being about halfway between 
the two. For the women the outstanding fact is the marked 
masculinity of the divorced group as compared with either 
the happy or the unhappy wives. 

It would be interesting to know why divorced women and 
unhappily married men tend to rate more masculine than the 
other groups with which they are compared. Analysis of the 
responses made by the three groups to the 420 items of the Strong 
occupational-interest test and to the 125 items of the Bernreuter 



120 SEX AND PERSONALITY 

personality inventory brings out a number of facts which are 
in line with the test results for the two groups in question. 1 
For example, the item responses of unhappily married men 
strongly suggest that this group is characterized by less tolerance, 
less sympathy, less amiability, less interest in uplift activities, 
and definitely less interest in such cultural things as symphony 
concerts, modern languages, philosophy, literature, and art. 
Nearly all of these characteristics would stamp them as masculine 
rather than feminine. The differentiating traits of the divorced 
women stand out even more clearly. The item responses of this 
group indicate that the typical divorced woman differs from both 
the happily and the unhappily married in being more self- 
assertive, more ambitious, less docile, more of an individualist, 
more intellectual, less moved by sympathy, and less interested in 
social welfare schemes; in short, her personality lacks the element 
of sweet femininity but may command respect for its masculine 
qualities of rugged strength and self-sufficiency. It is impossible 
to say to what extent these characteristics of divorced women 
are to be attributed to the selective effect of divorce itself, or to 
what extent they are the result of the broader contacts and 
outside employment which in the case of women are so likely 
to follow divorce. Perhaps both factors are involved. 

SUMMARY 

1. Masculinity and femininity as measured by the M-F test 
is definitely (though not closely) correlated with introversion- 
extroversion as measured both by the Laird and by the Conklin 
tests, masculinity being an extrovert trait, femininity an intro- 
vert trait. The relationship holds for both sexes at the college 
level, but is more marked for men than for women. In the case 
of delinquent boys no correlation was found between M-F scores 
and introversion-extroversion as measured by the Neyman- 
Kohlstedt test. 

1 JOHNSON, WINIFRED BENT, and LEWIS M. TERM AN, Personality characteris- 
tics of happily married, unhappily married, and divorced persons, Char, and 
Personal, 1935, 3, 290-311. 



CORRELATION OF M-F SCORE WITH PERSONALITY 121 

2. No reliable correlation was found between the M-F scores 
of delinquent boys and scores on the Cady and Raubenheimer 
battery of character tests. 

3. M-F scores are positively correlated in the case of males 
with scores on both the Stenquist and the McQuarrie tests of 
mechanical ability; this correlation does not hold for young 
women or girls. This is in agreement with the data on interests 
in Chapter IX. 

4. High-scholarship college men are more feminine, low- 
scholarship men more masculine. This is in contrast to the 
finding for college students in general that intelligence has a 
small positive correlation with masculinity (Table 34), The 
interests of high-scholarship men are more cultural, those of 
low-scholarship men more mechanical and athletic. Scholarship 
correlates negatively with M-F score for college women as well 
as for college men, but this trend is slight as compared with the 
large (opposite) influence of intelligence. Both trends are per- 
sistent and the results are in agreement with data presented in 
preceding chapters. 

5. The M-F scores of athletes tend to be strongly masculine. 
This is true of both sexes but is especially marked in the case of 
women. The mean M-F score of 55 outstanding women athletes 
was by far the most masculine found for any female group of this 
size. 

6. No reliable correlation was found between husband-wife 
resemblance or difference in M-F score (on a substitute test) and 
an index of their marital happiness. It appears, however, that 
men who are unhappily married tend to test reliably more 
masculine than divorced men, and that divorced women as a 
group test definitely more masculine than either happy or 
unhappy wives. 



CHAPTER VII 

RELATION OF M-F SCORE TO AGE, EDUCATION, 
AND INTELLIGENCE 

The M-F test consistently differentiates the sexes. Any pair 
of comparable sex groups which if combined form a single homo- 
geneous social unit always yield average scores that are widely 
divergent for the two sexes. Groups of men test on the average 
within well-defined limits on the masculine half of the M-F range 
and groups of women within a narrower range on the feminine 
half of the scale. However, there are often wide and significant 
differences between the means of groups within the same sex. 
In the present chapter we shall discuss some of the factors which 
influence these placements in the case of the large representative 
groups which make up the bulk of our test population, particu- 
larly the factors of age, education, and intelligence. 

THE INFLUENCE OF AGE 

Age and M-F score are found to be correlated in youth and in 
maturity, but the relationship, far from a close one at any 
period, is reversed in the later as compared with the earlier period 
of life. Intercomparison of the means of large and homogeneous 
groups from the early and middle adolescent years shows a 
fairly well-defined tendency in both sexes for masculinity to be 
correlated with increase in chronological age (see Figs. 5 and 6). 
In late adolescence or early maturity a change takes place in the 
direction of the trend so that as age increases the M-F score 
becomes less and less masculine until the close of life. 

There is a sex difference as to the age at which the reversal of 
the relationship occurs and also with respect to the amount of age 
influence. The mean score of males reaches its masculine peak 
somewhere during the high-school years and from this time 
onward through early and late maturity it declines steadily 

122 



RELATION OF M-F SCORE TO AGE 



123 



toward the feminine end of the scale, though no male group 
equals in femininity the corresponding female group (see Fig. 5). 
The mean of females continues to become more and more 
masculine from early adolescence to early maturity, the peak of 
masculinity being reached in the college years. From then on 



8th Uth Coll. Adults by Decades 
Grade Grade Soph. 20's 30's 40's 5p>__6Q>__78fc_ 



+ 80 
+ 70 
tbO 
t50 
+ 40 
+ 30 
+ 20 
+ 10 

-10 
-20 
-30 
-40 
-50 
-60 
-70 
-80 
-90 
.inn 




















/ 


^ x ^-* 
















/ 






"X 


a/a/es 










/ 








\ 










f 










\ 


















> 


\ 


















\ 




















^s 


-^ 
















































































































/ 


\ 
















/ 


> 


V--' 


~* Fen 


Tor/es 








x 


f 




> 


^^ 


^ 


p* ^ 


b- -* 


- 


X 



















80s 



FIG. 5. Age trends of mean M-F score. 

mental masculinity declines slightly and irregularly but unmis- 
takably until old age. 

In Fig. 5 it will be noted that the distance between the most 
masculine and the most feminine mean score is just twice as great 
for males as for females. In the case of males the difference is 
about 70 points, the distance between the means of high-school 
juniors and of eighty-year-old men. The most feminine mean 
for females, that for the eighth grade, is 35 points below the mean 
of college women, the most masculine female group. The mean 



124 SEX AND PERSONALITY 

score of men in old age is far below the mean position of the 
eighth-grade boys, but the final point reached by women in their 
later decades is not the lowest score norm. The eighth-grade 
girls are in terms of this test more feminine than the elderly 
women. From fifty years onward the women score lower on 
the scale than girls in the junior year of high school, but not so 
low as eighth-grade girls. Throughout this period their average 
score is some 25 points more feminine than the average college 
sophomore woman. 

The age trends indicated above may be described as we follow 
the curves from the point of view of the course of development 
of the average individual, boy or girl, man or woman. We may 
think of the endogenous predispositions and the exogenous 
influences that affect mental masculinity-femininity as operative 
continuously from birth onward. We have as yet no measures 
of the trait in early childhood. But as the average child enters 
adolescence he already registers upon the M-F scale within the 
range characteristic of his sex. From his last years in the 
grammar school onward until he reaches approximately his 
third year of high school the combined effect of nature and 
nurture makes the boy mentally more and more masculine. 
The increase in his masculinity from the fourteenth to the 
sixteenth or seventeenth year is considerable. During the later 
high-school years, however, the boy's score ceases to gain in 
masculinity and by the college sophomore year it registers in 
the feminine direction. The difference is not large, less than 
once its standard deviation, but decade by decade the modifica- 
tion continues. In the thirties the eighth-grade score point is 
passed. Each decade up to the sixties brings a larger drop than 
the decade preceding. In his seventies and eighties the average 
man in our test population has traversed about half of the 
distance that in the adolescent period divides the sexes in respect 
of mental masculinity and femininity. 

The course of the M-F trait in the average woman's life is 
rather different from this. In early adolescence she, like her 
brother, is becoming mentally more masculine. The difference 
between her scores at fourteen years and at sixteen or seventeen 



RELATION OF M-F SCORE TO AGE 125 

years is, however, considerably less than in the case of the boy. 
This is not mathematically a significant difference in our data. 
But after the third high-school year and if the girl goes on to 
college her score becomes increasingly masculine and the differ- 
ence between her M-F score in grammar or high school and 
in the college years is statistically significant. The average 
college girl registers 35 points more masculine on the M-F scale 
than the average eighth-grade girl, the difference being six 
times its standard deviation. In the same period the boy has 
gained 27 points but has lost 5 of them. By the time he is a 
college sophomore he is 22 points above the average eighth-grade 
boy. The latter difference is three times its standard deviation. 
The woman declines from her peak of mental masculinity 
reached during college days until in her sixties she is only 6 points 
more masculine than the average eighth-grade girl and is 29 
points below her masculine high point. At or near this position 
she tends to remain to the end of her life. 

The conspicuous contrasts between the sexes with respect 
to the influence of age upon M-F score are (1) in the extent of the 
initial rise from eighth grade to the third high-school year; (2) 
in the direction of the trend from the third high-school year 
to the second college year; (3) in the extent of change in score at 
that period; (4) in the age at which the turning point in develop- 
ment of the M-F trait occurs; (5) in the extent of decrease in 
masculinity from twenty years of age onward. It is noteworthy 
that at all ages and whatever the trends, in any comparable pair 
of groups the sexes retain their characteristic positions relative 
to each other. The scores for the sexes diverge more at the 
high-school than at the grammar-school age because, although 
both sexes have become mentally more masculine than before, 
the boys have gained more than the girls. The sex difference 
is less in the college period than in the high-school period because 
the men have become more feminine, the women more mascu- 
line. 1 From this age onward the sexes differ less and less 

1 The possibility of college acting as a selective factor should not be over- 
looked; perhaps the more feminine type of girl is more likely than the masculine 
type to abandon her education at this point with a view to early marriage, 



126 



SEX AND PERSONALITY 



because of the feminizing of the men's scores. Both sexes are 
becoming continually less masculine, but the change is more 
marked in men than in women. At eighty years of age they are 
closer together than at any other period. The distance here 
is 89 points as compared with 157 points at the high-school age. 
Statistical evidence of the relationship of M-F score to age is in 
the form of (1) comparisons of the means of age groups in terms 
of the ratio between their difference and the standard deviation 

TABLE 22. COMPARISON OF MEAN M-F SCORES or HOMOGENEOUS GROUPS AT 
DIFFERENT AGE LEVELS (MALES) 



Groups 


No. 
of 
cases 


Age 


M-F score 


Mean 
score 


S.D.M 


Diff. 


S.D. d iff. 


Eighth grade (unselected) 


95 
79 
130 

95 
208 
79 
344 
130 
531 

79 
45 
30 

124 
131 
44 

46 

53 

63 


14+ 
16+ 
20+ 

14+ 
20 to 90 
16+ 
20 to 90 
20+ 
20 to 90 

16+ 
18 to 25 
30 to 50 

15 
17+ 
26+ 

21 
26 

20 to 90 


+45.6 
+72.7 
+67.4 

+45.6 
+21.8 
+72.7 
+45.4 
+67.4 
+50.4 

+66.2 
+79.5 
+31.1 

+43.7 
+55.7 
+66.2 

+19.4 
+ 9.0 

+12.6 


5.42) 
5.25* 
4.18} 

,, 

3.40) 

,., 

2.92) 
4.18) 
2.05) 

5.28\ 
4.12( 
7.55} 

4.10) 
4.89; 
6.97) 

5.14) 

9.121 

4.52* 


-3.59) 
>-3.19 
+ .78/ 

+3.73 
+4.54 
+3.65 

;;> 

:;:}- 

!>- 


High-school juniors (unselected) 
College sophomores (unselected) . 

Eighth grade (unselected) 


Adults: Grade-school education 
(unselected) 


High-school juniors (unselected). 
Adults: High-school education 
(unselected) 


College sophomores (unselected). 
Adults: College education (unse- 
lected) 


Gifted boys 


Superior college students 


Who's Who men 


Younger delinquents 


Older delinquents 


Adult army prisoners 


Catholic student priests 


Protestant theological students . . 
Clergymen and ministers (prot- 
estant) 





RELATION OF M-F SCORE TO AGE 



127 



of the difference (Tables 22, 23, and 24) and (2) correlations 
between age and M-F score (Table 26). Both kinds of evidence 
are available for all parts of the curve. In connection with the 
correlation material an attempt has been made to isolate from 
the age factor certain other influences, such as those of mental 
age and marital experience. 

We shall present first the data for each sex derived from the 
comparison of means. The differences have been found between 
successive means from the youngest to the oldest groups and 
the ratios of these to their respective standard deviations have 
been computed. In general we may regard a difference as 
definitely significant and indicative of a clear trend when it is 

TABLE 23. COMPARISON OF THE MEAN M-F SCORES OF HOMOGENEOUS GROUPS 
AT DIFFERENT AGE LEVELS (FEMALES) 



Groups 


No. 
of 
cases 


Age 


M-F score 


Mean 
score 


S.D. M 


Diff. 


S.D.diff. 


Eighth grade 


84 
86 
130 

84 
287 

86 
820 

130 
760 

81 
92 
25 

78 
64 

50 
41 


14+ 
16+ 
20+ 

14+ 
20 to 90 

16+ 
20 to 90 

20+ 
20 to 90 

16+ 
20+ 
30 to 50 

21 + 
20 to 90 

18 to 25 
20 to 90 


-95 3 
-84 8 
-60 8 

-95.3 
-86.2 

-84 8 
-84 7 

-60 8 
-74.7 

-57 3 
-36 3 
-45 5 

-63.3 
-64 4 

-71.3 
-85.8 


4.62} 

A. 77 

3A3} 

4 62? 
2.53} 

477? 
1.47} 

3.43? 
1.60} 

5 10} 
4 38f 
8.47; 

4.47? 
5.27} 

5.60? 
5.27} 


- -n.-s.99 

-4.08J 
-1.72 
- .02 

+3.68 

-3.13^ 
+2.12J 1>19 

+ .15 
+1.89 


High-school juniors. ... 


College sophomores 


Eighth grade 


Adults: Grade-school education 
High-school juniors 


Adults: high-school education 
College sophomores 


Adults: College education. . . 
Gifted girls 


Superior college students 


Who's Who women 


Student nurses . . 




Music students 


Musicians and music teachers . . . 



128 



SEX AND PERSONALITY 



as great as three times its standard deviation. When the 
difference continues to be even twice its standard deviation in 
successive groupings a trend is fairly definitely indicated. We 

TABLE 24. AGE DIFFERENCES IN M-F SCORES OF ADULTS 



Male 


Female 


Single 
decades 


No. 
of 
cases 


Mean 


S.D.M 


Diff. 


No. 
of 


Mean 


S.D.M 


Diff. 


S.D.diff. 


S.D. d iff. 




















20's 


342 


+ 57 9 


2 76^ 




604 


- 74.2 


1 82^ 




30's 


330 


+ 49 5 


} 

305 


+ 2 04 
}+2 97 

+2.05; 


488 


- 84 5 


" 


+ 3 90 
> + l 36 
-1.9l' 


40's 


178 


+ 39 7 


f * J 


J+4.24 


328 


- 78.4 


2 54 


}+ -64 


50's 


108 


+ 26.6 


*J 


+ 2 28; 
}+4 68 

+ 2 37; 


234 


- 86.5 


,'J 


+ 2 28^ 
+ .66^ 


60's 


75 


+ 10 8 


4 99^ 


f+2.81 


153 


- 89 1 


y 
~\ 










| 


+ .75^ 






} 


55 


70's 


38 


+ 3 7 


8 03^ 


}+ .87 


100 


- 86 5 


3.70 7 


} 43 








} 


+ 26* 






} 


+ 21 


80's 


6 


+ .5 


9 13* 




18 


- 88 4 


8 24' 




90's 










1 


-121 






Combined 


















decades 


















20's& 


672 


+ 53.8 


2 06\ 




1092 


- 78 8 


1 34-v 




30 f s 
























> 


+ 2.45\ 






> 


+ 1 60\ 


30's & 






) 


1 






( 




40's 


508 


+ 46 




>+5 39 


816 


- 82 


1 H> 


> + l 32 


40's& 
50's 


286 


+ 34 8 


2 857 


+ 2.04* 


562 


- 81 8 


} 


- .J 


50's& 
60's 

60's& 
70's 


183 
113 


+ 20.1 

+ 8.4 


3 39/ 
3 60/ 


+ 3 31) 
>+5 75 

+ 2 37) 


387 
253 


- 87 5 
- 88 1 


: 

2 34/ 


+ 2 17j 

>+2 13 
+ .18) 


70's& 
80's 


44 


+ 3 2 


7 04/ 


+ .66 


119 


- 87 2 


J 


- 23 



have made combinations in our groupings where the trend was 
suggested in this way, continuing the combining process until 
the evidence was convincing either for or against the apparent 
tendency. 



RELATION OF M-F SCORE TO AGE 129 

Our analysis proceeds from the youngest groups to the oldest. 
The first step shows an increase of 27.1 points in score between 
eighth-grade and high-school junior boys, or 3.59 SJ).^ 
(Table 22). The next step, a drop of 5.3 points from the high- 
school norm to the college sophomores is less than its S.D. and 
therefore only meaningful in connection with other results. It is 
worthy of note that the pronounced change in score before age 
twenty-five occurs between the eighth grade and the junior year 
of high school and that the former group is significantly less 
masculine as compared with the latter and also as compared 
with the college norm. High-school juniors and college sopho- 
mores, however, are not significantly divergent. 

From the turning point, the mean score of the high-school 
junior boys, the line of direction is toward the feminine end 
of the scale. Pursuing the course of the total population group 
(Table 24) we find each successive decade position less masculine 
than the preceding one. The difference between the means 
for the high-school junior boys and the twenty-year-old men is 
almost twice (1.9) its S.D. The means for succeeding decades 
up to the sixties differ by something over 2.0 S.D.^ff.. From the 
sixties to the seventies the curve still falls in the feminine direc- 
tion, and similarly from the seventies to the eighties, but in these 
cases with their small populations the differences are less than 
their respective S.D.'s. Enlarging populations by combining 
groups anywhere along the line gives statistically reliable differ- 
ences between stages. It is to be noted that we are following 
here the total population group only. The separate groups 
of adults representing three educational levels will be considered 
later. 

Turning now to the norms for the girls and women we find a 
similar but statistically less significant tendency. The first 
step, from the eighth grade to the high-school-junior level 
(Table 23) is not statistically reliable. The second step, from 
the high-school-junior to the college sophomore level, 24.0 points, 
is more than 4 5.0.^., whereas the corresponding step for the 
boys is not significant statistically and is in the opposite direction. 
The step from college sophomores to twenty-year-old women 



130 



SEX AND PERSONAUT7 



combining all degrees of education is 13.4 M-F points or 3.45 
S.D.diff.. In the general population group (Table 24) that from 

TABLE 25. COMPARISON OF MEAN SCORES OF MALE AND FEMALE GROUPS AT 
SUCCESSIVE AGE PERIODS 





TV/Tpori 


S D M 


Diff 


Diff. 










S.D.diff. 


School and college groups: 
14J years (Eighth-grade pupils) 
Male 


+45.6 


5.42) 






Female 


-95.3 


4.62} 


+ 140.9 


+ 19.8 


16>6 years (High-school junior 
pupils) 
Male 


4-72.7 


S 25| 






Female 


-84.8 


47 7 f 


+157.5 


+22.2 


20)^ years (College sophomore 
students) 
Male 


+67 4 


4 18) 






Female ... 


-60.8 


3.43} 


+ 128.2 


+23.7 


Decade groups of adults: 
20's 
Male 


+57 9 


2.76) 






Female . . 


-74 2 


1 82} 


+ 132 


+39.9 


30's 
Male 


+49 5 


3 05) 






Female 


-84 5 


1.93} 


+ 134 


+37.1 


40's 
Male 


+39 7 


3 64) 






Female 


-78 4 


2 54J 


+ 118 1 


' +26 6 


SO's 
Male. 


+26 6 


4.46) 






Female 


-86 5 


2 50} 


+ 113 1 


+22 1 


60's 
Male 


+10 8 


4 99) 






Female 


-89 1 


3 02J 


+99 9 


+ 17.1 


70's 
Male 


+ 3.7 


8 03) 






Female 


-86 5 


3.70} 


+90 2 


+ 10 2 


80's 
Male 


+ 0.5 


9.13) 






Fern**)* 1 -- - - - - 


-88.4 


8. 24J 


+89 


+ 7.2 













the twenties to the thirties is 3.9 S-D.^. The step from the 
thirties to the forties is in the reverse direction and is almost 



RELATION OF M-F SCORE TO AGE 131 

2 S.D.aiff.. From the forties to the fifties, the trend is again 
feminine and the distance is 2.3 S-D.^. From the fifties to the 
sixties the feminine tendency continues but the difference is less 
than its standard error. The remaining steps, one upward from 
the sixties to the seventies and one downward from the seventies 
to the eighties, are small and unreliable. Irregularity is obviously 
a feature of the course of the women's decade norms. Com- 
bination into successively larger groups, however, shows differ- 
ences between younger, middle, and older adult groups that are 
at least twice their respective S.D.'s; and comparison of persons 
under fifty with those over sixty years 4.7 its S.D. 

The two most important differences between the curves for the 
sexes are again emphasized by the statistical treatment: (1) 
the difference in the age level at which the curves turn from their 
early masculine to a feminine direction, and (2) the range from 
the most masculine to the most feminine score in the life of 
each sex. The latter sex difference is emphasized statistically 
by the fact that the successive steps in the male curve are so 
much greater, so much more consistently in the same direction, 
and so much more reliable. 

The differences between corresponding sex groups are indicated 
in Table 25. These show the three younger sex groups differing 
by 128 to 157 M-F points; 20 or more times the standard errors. 
The adults compared by decades show the largest sex differences 
in the twenties and thirties (over 130 M-F points with critical 
ratios of 37 and 40), but all the sex differences are significant even 
among the older, smaller populations. Those at the forties and 
fifties are more than 110 M-F points; the sixties, the seventies, 
and the eighties have differences of 89 to 100 M-F points. In 
spite of the modifying influence of age the M-F scale continues 
to differentiate the central tendencies of the two sexes from 
youth to old age. 

We may turn now to the correlational evidence bearing on the 
relationship between age and M-F score (Table 26). Correla- 
tions are available for six groups of boys and for seven groups of 
girls, together with a correlation for each sex from two of the 
respective groups combined into a single population. The corre- 



132 



SEX AND PERSONALITY 



lations in this age range are all quite small. The coefficients 
for the boys include only one significant value and it is for a very 
small and probably not unselected population, the orphans. 
Eighth-grade boys, high-school juniors, gifted boys, younger and 

TABLE 26. CORRELATIONS BETWEEN M-F SCORES AND CHRONOLOGICAL AGE 
IN YOUNGER POPULATIONS 



Male groups 


No. 
of 
cases 


Mean 
C.A. 
Yr. 
mo. 


Age range 


Correla- 
tion 
between 
C.A. and 
M-F scores 


Correla- 
tion 
between 
C.A. and 
M-F score 
with M.A. 
constant 


Correla- 
tion 
between 
M-F score 
and M.A. 
with C.A. 
constant 


Orphans 


15 


14-2 


11-11 to 16-7 


+ .601 .12 


+ .551 13 


- 201 16 


Eighth grade ... . 
High-school juniors. 
Combined grade and 
high school . 


95 
79 

173 


14-4 

16-8>j 

15-4 


12 to 17 
14 to 21 

12 to 21 


+ 03 .07 
- 03 .08 

+ 14 06 


+ .081 05 


+ .151 05 












Correla- 
tion 
between 
C.A. and 
M-F score 
with edu- 
cational 
placement 
constant 


Correla- 
tion 
between 
educa- 
tional 
placement 
and M-F 
score with 
C.A. 
constant 


Gifted boys 
Young delinquents 


79 
124 


16-9 
15 


11 to 20 
12 to 17 


+ .08 08 
+ 02 06 


+ .121 .08 
001 06 


00+08 

+ 051 06 


Older delinquents. . . 
Superior college men . . 
Superior college men with 
high scholarship 
Superior college men with 
low scholarship 


131 
97 

46 
46 


17-6 
21 

21 
21 


10 to 20 
18 to 25 

19 to 25 
19 to 25 


+ .05 06 
.001 .07 

-.15 10 
-.13 10 


+ 041 06 


+ 131 06 


Female groups 










Correla- 
tion 
between 
C.A. and 
M-F score 
with M A. 
constant 


Correla- 
tion 
between 
M-F score 
and M.A. 
with C.A. 
constant 


Orphans . . . 
Eighth grade. . . 
High-school juniors 
Combined grade and 
high schoof . . . 


13 
84 
86 

170 


14-9 
14-2 
16-5 

15-3 


12-1 to 17-2 
12 to 19 
14 to 21 

12 to 21 


+ 39 16 
+ 28 07 
- 08 .07 

+ 16 05 


+ 291 .07 
+ 071 05 


+ 101 07 
4 141 05 












Correla- 
tion 
between 
C.A. and 
M-F score 
with 
educational 
placement 
constant 


Correla- 
tion 
between 
educational 
placement 
and M-F 
score with 
C.A. 
constant 


Private school. . . . 
Gifted girls 
Superior college women . 
Student nurses 


28 
81 
91 
78 


15-11 
16-2 
20-2 
21-1H 


13 to 19 
13 to 19 
18 to 24 
18 to 25 


+ .31.ll 
+ .20 .07 
-.031 08 
-.151 08 


+ 021 13 
+ 101 07 


+ 351 11 
+ 051 07 



RELATION OF J/-F $COR TO AGE 13$ 

older delinquents, and even the combined grade and high-school 
groups show correlations that are with one exception no larger 
than their probable errors. 

For the girls also the coefficients are low, though not generally 
so low as for the boys. Two of them are three times the P.E. 
and two others almost reach this criterion of reliability. The 
correlations between M-F score and age seem from these figures 
to be significant in the eighth grade, and in the combined eighth- 
grade and high-school group. They appear to be nearly so 
for the orphans, the private-school girls, and the gifted girls. 
Apparently a slight but fairly persistent relationship is indicated 
in the case of the girls, whereas for the boys there is barely a 
suggestion of relationship. 

The single significant coefficient in the case of the boys is that 
for the small group of orphans. The mean M-F score of this 
group is so unusual for this age in its intense masculinity as to 
suggest that many of the individuals in the group may have been 
responding to a special social influence at the time the test was 
taken. Several of them, especially the older ones, were antago- 
nistic to the management of their Home at the time the test was 
given and behaved in an aggressively hostile manner during 
the test. This might account not only for the unusually mascu- 
line mean score but also for the correlation with chronological 
age. For this reason and also because of the small size of the 
group the mean and the correlation can be given little weight. 
The only other correlation of any consequence among the boys' 
groups is that for the combined eighth-grade and high-school 
groups. The coefficient, .14 .06, although small is of interest 
because the population in question represents the most consider- 
able age range analyzed in this connection by the correlation 
method. We have probed its possible significance by deriving 
the correlation between age and M-F score with mental age 
constant and find it to be negligible, +.08 .05. The relation, 
however, between M-F score and mental age, with age constant, 
is +.15 .05, indicating a small but probably true correlation. 
The negligible correlation for the high-school-junior group is to be 
expected, as the trend of development of the M-F score is altered 



134 SEX AND PERSONALITY 

at about this age. Among college men the relationship between 
age and M-F score is also negligible. 

More detailed analysis of the coefficients for the girls' groups 
gives evidence of an age correlation, but does not make clear 
whether this is due to age alone or to some related factor. In the 
eighth-grade group the relation between age and M-F score 
appears to be a genuine one, not attributable to the mental-age 
factor. In the combined eighth-grade and high-school group, 
however, mental age appears to contribute the larger influence. 
Thus it seems that the relationship has shifted between the 
eighth grade and the third high-school year. Among private- 
school girls and gifted girls there is a suggestion of correlation 
between age and M-F score. This is found for the first named 
to be chiefly associated with grade placement, perhaps a measure 
of mental age; for the latter also mental age is probably account- 
able. Among the older school groups of girls no correlation 
between age and M F score appears. This is, of course, to be 
expected in view of the fact that the direction of the trend in 
M-F score is turning among the girls at about the age represented 
by the groups in question. 

Summarizing the correlational data for the younger groups, we 
find that the relationship between age and M-F score is positive 
but very slight in the regular school groups of adolescent boys 
living under normal home conditions, and that this relationship 
is probably largely a matter of mental age. Among older 
adolescent boys the relationship is negligible, and in the college 
years also it is not demonstrable. The M-F scores of girls 
in the teens apparently correlate slightly with age. It is not 
clear that this is attributable to the influence of mental age, 
though there is some indication that mental age is involved. 
With college girls there does not seem to be any clear relationship 
between M-F score and age. 

In addition to the correlations between age and M-F score 
for the younger groups, coefficients between age and M-F score 
have also been derived for both sexes in the populations of adults 
who had been married and who reported having had all or part 
of a high-school education. The group in question is called the 



RELATION OF M-F SCORE TO AGE 135 

high-school adult group. As was to be expected from the com- 
parison of age means, significant correlations were found for both 
sexes, though the coefficient for the women's group was exceed- 
ingly small. They are: for the men, r = .26 .04; for the 
women, r = .09 + .02. The scores of the unmarried members 
of the same educational group were also correlated with age. 
The coefficients for these were as follows: for the men, 
r = -.12 .09; for the women, r = -.11 .06. It should be 
mentioned that in the case of the single men the age range was 
much narrowed. There was only one unmarried man over 
thirty-one years of age, but the ages of the single women extended 
from twenty to eighty years of age as did those of both married 
groups. 

In order to test whether married life itself was the chief factor 
in the feminine trend with increasing age, the correlations 
between M-F score and the number of years married were 
calculated. They were found to be as follows: for the men, 
r = -.28 .05; for the women, r = -.07 .03. Partial 
coefficients were then derived between age and M-F score 
with the number of years of married life constant. These 
resulted as follows: for the men, r = .02 .05 ; for the women, 
r = .06 .03. The converse coefficients (M-F score corre- 
lated with the number of years of married life, age being con- 
stant) were as follows: for the men, r = .11 .05; for the 
women, r = +.01 .04. 

Reviewing the correlations between age and M-F score among 
married and unmarried adults of high-school education and the 
correlations between number of years married and M-F score 
in the married part of the group, it appears that, among the 
married at least, age and M-F score are somewhat correlated 
negatively in both sexes, but that in the case of the women the 
relationship is very slight. As far as married life is concerned, 
while for the men the number of years thus spent is slightly but 
significantly correlated negatively with mental masculinity there 
is no significant correlation either positive or negative in the case 
of the women. When the element of age is made constant the 
correlation with number of years married disappears for the men. 



136 SEX AND PERSONALITY 

Summarizing all of the data with respect to age and M-F score, 
we find that there is a demonstrable relationship between these 
two in both sexes from the earliest to the latest age for which 
scores are available. We have discovered further that the 
direction of the influence changes, during adolescence for the men, 
in early maturity for the women, and that whereas before the 
change age correlated positively with M-F scores, after the 
change the correlation is a negative one. Possibly mental age is 
involved in the positive correlation in childhood and also in 
the negative correlation in later maturity. It is suggested by the 
data that the correlation with age is greater for the younger girls 
than for the younger boys. The correlation in later years is 
greater for the men than for the women. 

In childhood and youth both boys and girls become mentally 
more and more masculine as they proceed through the later 
grammar grades and high school. Thereafter the boys become 
mentally more feminine, while the girls wax increasingly mascu- 
line until some time during the college years. Then they too 
begin gradually to lose their masculinity, although unlike the 
men they never again become quite so feminine mentally as 
they were before they left the grammar school. In contrast to 
this the men are on the average never again after their thirties as 
masculine in this test as they were in the eighth grade, in high 
school, or in college. 

The most striking contrast between the sexes is in the greater 
modification of the men from decade to decade, as compared 
to the slight alteration in the women during the same periods. 
The limits between which the mean scores of females register 
are set by the beginning and the end of adolescence. Those for 
males are set by middle adolescence and late senescence. Appar- 
ently man draws nearer and nearer to woman's norm, whereas 
woman deviates little from her fundamental position. 

Finally it seems important in view of the changes that do take 
place to reiterate that in no case does the mean of either sex pass 
over the mid-line between the sexes. At no age do corresponding 
groups from the two sexes ever approach each other in mean 



RELATION OF M-F SCORE TO AGE 



137 



M-F score, although individuals of either sex may score far 
within the range of the other. 

TABLE 27. SCOHE DISTRIBUTIONS OF ADULT POPULATION BY AGE DECADES 





Age 


Score 


20-29 


30-39 


40-49 


50-59 


60-69 


70-89 


Total 




M 


F 


M 


F 


M 


F 


M 


F 


M 


F 


M 


F 


M 


F 


221- 240 






























201- 220 






























181- 200 


3 




2 




















5 




161- 180 


2 
























2 




141- 160 


10 




7 




3 
















20 




121- 140 


19 




11 




6 




3 












39 




101- 120 


37 




28 




13 


1 


3 




2 








83 


1 


81- 100 


41 


2 


44 




14 




5 




3 




1 




108 


2 


61- 80 


64 


3 


42 




22 


3 


14 




3 




3 




148 


6 


41- 60 


43 


3 


46 


1 


29 




15 


1 


9 




5 




147 


5 


21- 40 


49 


8 


45 


1 


28 


2 


23 


1 


16 




2 




163 


12 


1- 20 


30 


11 


40 


10 


25 


6 


15 


1 


12 


3 


4 




126 


31 


19 


21 


39 


20 


16 


18 


18 


14 


6 


10 


4 


10 


2 


93 


85 


- 20 39 


14 


56 


15 


44 


13 


34 


7 


19 


8 


4 


4 


9 


61 


166 


- 40 59 


3 


97 


7 


70 


4 


43 


6 


27 


5 


20 


2 


11 


27 


268 


- 60 79 


1 


100 


3 


74 


2 


55 


2 


44 


3 


27 


2 


17 


13 


317 


- 80 99 


4 


106 


3 


91 




58 




44 


1 


35 




24 


8 


358 


-100- -119 


1 


90 




79 


1 


52 




51 




28 




15 


2 


315 


-120 139 




53 




60 




32 


1 


19 




15 




22 


1 


201 


-140- -159 




25 




26 




13 




18 




11 


1 


5 


1 


98 


-160 179 




6 




8 




9 




3 




4 




2 




32 


-180- -199 




4 




7 




1 
















12 


-200 219 




1 








1 
















2 


-220 239 








1 




















1 


M 


58 16 


48 75 


39 71 


26 61 


12 44 


1 68 


44 27 


Mean 


-74 17 


-84 50 


-77 98 


-85 91 


-88 57 


-91 00 


-81 02 


M 


51 15 


51 27 


48 53 


46 31 


43 23 


48 31 


51 68 


* ' F 


45 19 


42 61 


45 91 


38 31 


37 82 


36 35 


43 22 


M 


342 


313 


178 


108 


72 


34 


1047 


N F 


604 


488 


328 


234 


151 


107 


1912 



Decade comparisons in Chapter VII and references in other chapters to the means for 
the total population show slightly different values from those in this table. The discrep- 
ancy, in no case significant, is due to the exclusion or inclusion of small special groups. 

In Table 27 are distributions of scores for unselected adults 
of each sex by decades from the twenties through the eighties. 



138 SEX AND PERSONALITY 

THE INFLUENCE OF EDUCATION 

The course of mean M-F scores of typical groups has been 
followed from youth to old age. We have seen it in both sexes 
rise in the early years toward a masculine peak and then fall 
gradually, as age advances, toward a feminine limit. In viewing 
the changes from grammar school to high school and from high 
school to college the question naturally arises whether in the 
mature groups te M-F score might be found correlated with 
education as well as with age. Only a few correlations are 
available on this point. The chief evidence is contained in a 
comparison of the means at given ages for groups of varying 
amounts of education. The means for the younger school 
groups range themselves according to the level of education 
reached. We have subdivided the adult population of each 
sex into three groups: (1) persons whose education went no 
further than the eighth grade, (2) those who had all or part of a 
high-school education or its equivalent, and (3) those with all 
or part of a college course. Examination of the curves for each 
of these three educational levels shows that they tend severally 
to follow the course of the group as a whole. Comparing the 
means for these groups we have computed the differences 
between them at successive ages and the reliability of the 
differences. 

In Fig. 6 we observe certain striking and persistent features in 
the mean M-F scores at the three educational levels. For the 
men the differences between the adult populations of high-school 
and of college education are generally insignificant; the scores 
of the grammar-school group show a decided tendency to be 
persistently and, for combined groups, significantly more 
feminine than the scores of high-school or college men. In 
the case of the women, the adults of high-school and of grammar- 
school education do not differ significantly except at the seventh 
decade, but the women of college education diverge from both the 
other educational groups regularly and significantly. An 
attempt to interpret this sex difference may make it more 
meaningful. 



RELATION OF M-F SCORE TO AGE 



139 



It appears that the greatest masculinizing influence in boys 
(1) is present and becomes operative at this time in those indi- 
viduals who pursue their education beyond the eighth grade 
to the high-school-junior level or (2) is developed in them 
through exogenous pressures at this period. Whether inherent 



+ 70 
+ 60 
+ 50 
+40 
+ 30 
+ 20 
+ 10 


-10 
-20 
-30 
-40 
-50 
-50 
-70 
-80 
-90 
-100 



Adults by Decades 
,2QL30s40s_50's 60s 70s&80s 




College Ed. 
High school Ed. 
-Grade school Ed. 



Femoiles 




FIG. 6. Age trends of mean M-F score in relation to schooling. 



or induced, the mental masculinizing that occurs at this period 
persists in its expression in terms of the M-F score throughout 
life. Somewhat similarly in the case of the girls, the effect 
appears chiefly in those who pursue part or all of a college course. 
It appears most pronouncedly at the period between the third 
high-school year and the second college year and its effect as 
expressed in the M-F score persists thereafter throughout life. 



140 



SEX AND PERSONALITY 



Girls who leave school at or before the dose of the grammar 
grades and even those who continue through part or all of the 
high-school years do not at any age average as masculine as 
college women. 

Statistical evidence for the educational differences consists 
chiefly in comparisons between the averages. A few correlations 
between M-F scores and educational status are available (Table 
28) but these are usually of slight interest in the present connec- 
tion because they are for atypical groups and for short ranges. 
So far as they go the correlations offer no contradiction to the 

TABLE 28. CORRELATIONS BETWEEN M-F SCORE AND SCHOOL GRADE REACHED 



Groups 


No. of 
cases 


Correlation 
between M-F 
score and school 
grade 


Correlation 
between M-F 
score and school 
grade with C.A. 
constant 


Boys' groups: 
Young delinquents 


124 


+ 05 .06 


+ 05 06 


Older delinquents 


131 


+ 14 .06 


+ 13 06 


Gifted boys 


74 


+ .01 .08 


.00 .08 


Men 25 to 29 years old 


126 


+ .10 .06 




Girls' groups: 
Private-school girls 
Gifted girls . . .... 


28 
80 


+ .45 .11 
+ .18 .07 


+ .35 .11 
-.05 .07 


Women 25 to 29 years old . 


175 


+ .27 .05 





data already given. The coefficients for both the delinquent 
groups and for the two gifted populations are positive but small. 
They and possibly the coefficients for the private-school group 
may be disregarded since they are based on nonrepresentative 
groups and offer no contradiction to the manifest trends. The 
correlations for the two young adult populations are probably 
worthy of consideration since they represent the scores of 
persons who are just entering maturity, who have generally 
reached the end of their educational training, and who are drawn 
from all three educational levels. For the men the correlation 
is + -10 .06, which is not significant. For the women, how- 



RELATION OF M-F SCORE TO AGE 



141 



ever, r = +.27 .05, suggesting a small but probably actual 
relationship. 

We may consider next the mean scores of certain school groups 
ranging from the preadolescent to the early adult years (Tables 22 
and 23). The eighth-grade boys, as we have seen, are definitely 
more feminine than high-school-junior boys or college sopho- 
mores. Both differences are more than 3 S.D.ajff.. Unfortu- 
nately for purposes of calculation the comparison shows age as 
well as education as a variable. Young delinquents whose age is 
intermediate between the eighth-grade boys and the high-school 

TABLE 29. COMPARISON OF M-F SCORES OF ADULT POPULATION AT THREE 
EDUCATIONAL LEVELS (DECADE PERIODS) 





Male 


Female 


Group 












M 










No. 
t 


A/Too n 


Q Tl 


Diff. 


No. 
t 




o T\ __ 


Diff. 




OI 

cases 


Mean 


t> 1J.M 


S.D. d ,ff 


Ql 

cases 


can 


o.JJ.M 


S.D.diff 


20V. 




















Grade . . 
High school 
College . 


21 
117 
204 


+ 31 4 
+ 56 6 
+ 61 3 


10 56) 
5 07{ 
3 36) 


-2 15) 
[ -2 69 
- .76$ 


28 
257 
319 


-83 1 
-78 1 
-70.2 


8 63 
2 76 
2 55 




- .54) 
f-1.43 
-2.12$ 


30's: 




















Grade 
High school 
College 


46 
113 
164 


+ 19 2 

+ 52 1 
+ 55 9 


7 08 
5 06 

3 55 




-3 78) 
[ -4 63 
- 61$ 


54 
208 
204 


-88 
-88 2 
-79 


6 28 
2 96 
2 88 


\ 
\ 


- .02) 
} -1 30 
-2 21$ 


40V. 




















Grade 
High school 
College. 


44 
44 

83 


+ 36 
+ 43 2 
+ 39 9 


7 70 
7 39 
5 16 




- 68) 
[ - .43 
+ .37$ 


70 
118 
109 


-83 2 
-80 
-73 


5 76 
4 00 
4 71 




- 59) 
-1.37 
-1 14$ 


50's: 




















Grade 
High school 
College. . 


55 
27 
46 


+ 25 8 
+ 24 6 
+ 31.4 


5 93 
7 97 
7 38 




+ 12) 
[ - .59 
- .63$ 


52 
110 
72 


-89 5 
-89 9 
-79 2 


5 22 
3.41 
4.84 


; 


+ 06) 
[ -1.44 
-1 80$ 


60V 




















Grade . . 
High school 
College . 


23 
29 
18 


+ 1 9 

+ 18 8 
+ 11.6 


9 55 
6 86 
9 23 




-1 43) 
[- .73 
+ .62$ 


48 
74 
31 


-82.8 
-95.4 
-83 7 


5 31 
4 17 
6 90 




+ 1 87) 
[- .10 
-1.46) 


70's and 80' s: 




















Grade 
High school 
College .. 


14 
14 
16 


+ 1 9 
9 

+ 80 


9 19} 
17 46J 
8 64} 


+ 14) 
[- .48 
- .46$ 


37 
57 
25 


-90 
-87 
-80 7 


6 19 
4 91 
5 99 


! 


- .38) 
[ +1 08 
- .82$ 



142 



SEX AND PERSONALITY 



juniors but whose average school grade is much nearer the 
former than the latter, perhaps even below the former in terms of 
accomplishment, rate in M-F score not more masculine, but a 
shade more feminine than the eighth-grade boys. The older 
delinquents, whose chronological age is beyond that of the high- 
school juniors but whose educational status is near the eighth- 
grade level, are 17 M-F points below the high-school norm; a 
difference that is 2.4 S-D.^. When the two delinquent groups 
are combined their average age approximates that of the high- 
school-junior group, although their educational standing, cer- 

TABLE 30. COMPARISON OF M-F SCORES OF ADULT POPULATION AT THREE 
EDUCATIONAL LEVELS (TWO-DECADE PERIODS) 





Male 


Female 


Group 




















No. 






Diff. 


No. 






Diff. 




of 


Mean 


S.D.M 




of 


Mean 


S.D.M 
















S.D.<Uff 












cases 








cases 








20' s and 30's: 


















Grade. 
High school 
College. 


67 
230 
368 


+ 23 
+ 54 4 
+ 58 9 


5 92) 
3.49J 
2 45} 


-4 57) 
[ -5 59 
-1 05) 


82 
465 
523 


-86 3 
-82 6 
-73 6 


5 09) 
2 03] 
1.93) 


- .68) 
f-2.33 
-3 21) 


30's and 40 's: 


















Grade . . . 


90 


+ 27 4 


5 29 




-3 29) 


124 


-85 3 


4 26) 


- 03) 


High school 


157 


+49 6 


4 20 




X -3 81 


326 


-85.5 


2 36{ 


\ -1.70 


College. . 


247 


+ 50 5 


2 96 




- 17) 


313 


-76 9 


2 SO/ 


-2 49) 


40's and 50's: 


















Grade 
High school 
College 


99 
71 
129 


+ 30 3 
+ 36 1 
+ 36 9 


4 78 
5 59 
4 25 




- 79) 
V -1 03 
- 10} 


122 
228 
181 


-85 9 
-84 8 
-75 5 


4 00 
2 67 
3 43 




- 23) 
[ -1 98 
-2 49) 


50'sand 60's: 


















Grade 
High school 
College 


83 
56 
64 


+ 17 7 
+ 21 6 
+ 25 8 


5 23 
4 49 
6 01 




- 56) 
I -1.01 
- 57) 


100 
184 
103 


-86 3 
-92 1 
-80 6 


3.74 
2 65 
3 98 




+ 1 27) 
[ - .87 
-1 90i 


60'sand 70's: 


















Grade 
High school 
College . . 


39 
42 
32 


+ 2 3 
+ 11 9 
+ 11 1 


6 57) 
5 27 
6.70J 


-1 14) 

r- - 94 

+ 09) 


79 
121 

53 


-86 
-92 3 
-81 6 


4 22 
3 40 
4 78 




+ 1 17) 
[ - .69 
-1.83) 


70'sand80's: 


















Grade 
High school 
College 


14 
14 
16 


+ 1 9 
- 9 
+ 80 


9 19 
17 46 
8 64 




+ 14) 

c- 48 

- .46) 


37 
57 
25 


-90 
-87.0 
-80 7 


6.19 
4 91 
5 99 




- .38) 
V -1.08 
- .82) 



RELATION OF M-F SCORE TO AGE 



143 



tainly in terms of achievement, is not above the eighth-grade 
level. The difference between the two groups of equal age 
but of divergent educational status is 23 scale points and con- 
siderably more than 3 SJ).^.. Similarly the gifted boys, 
younger than the older delinquents, but averaging four years 
beyond them in educational advancement, score several points 
above them. That they do not score so far or so significantly 
above them as the high-school juniors is probably, at least in 
part, to be attributed to the fact that in the gifted boys the turn- 
ing point associated with greater maturity has already been 

TABLE 31. COMPARISON OF M-F SCORES OF ADULT POPULATION AT THREE 
EDUCATIONAL LEVELS (THREE-DECADE PERIODS) 





Male 


Female 


Group 
























No 








Diff. 


No. 








Diff. 




of 
cases 


Mean 


S.D.M 




of 
cases 


Mean 


S.D.M 




S.D.diff 


S.D. dif f 


20's. 30's. 40's: 






















Grade 
High school 
College 


111 
274 
451 


+ 28 2 
-1-52 6 
+ 55 4 


4 74 
3 24 
2 24 




-4 26) 
V -5 19 
- .70} 


152 
583 
632 


-84 9 
-82 1 
-73 5 


3 82 
1 81 
1 79 




- 66) 
[ -2 69 
-3 36) 


40's, 50's, 60V 






















Grade 
High school 
College 


127 
100 
147 


+ 24 
+ 31 1 
+ 33 8 


4 40 
4 51 
3 96 




-1 12) 
[ -1 64 
- 45} 


170 
302 
212 


-85 
-87 4 
-76 7 


3 24 
2 27 
3 11 




+ 59) 
> -1 86 
-2 78} 


60's, 70's, 80's: 






















Grade 
High school 
College 


42 
43 
34 


+ 1.9 

+ 12 4 
+ 99 


6 42 
7 47 
6 36 




-1 06) 

[ - 88 
+ 25} 


85 
131 
56 


-86 
-91 8 

-82 4 


4 05 
3 20 
4 67 


i 


+ 1 13) 
[ - .58 
-1 67) 



passed and in terms of mental maturity they should perhaps be 
compared in M-F standing with college sophomores rather than 
with high-school students of equal age. That there are other 
trends operative will be brought out in succeeding chapters. 

From the twenties onward the decade norms for the three 
educational levels in the above tables show persistent and fairly 
regular divergence. In the twenties and thirties the norms of 
adults of grade-school education appear to diverge significantly 
from the high-school and college norms. Norms for combined 



144 



SEX AND PERSONALITY 



decades show the differences to be reliable in the earlier decades 
and the trends to be persistent thereafter. Finally three com- 
bined groups including respectively the adult populations at the 
three educational levels show differences with large critical 
ratios. 

TABLE 32. COMPARISON OF M-F SCORES OF ADULT POPULATION AT THREE 
EDUCATIONAL LEVELS (ALL AGES COMBINED) 



Group 


Male 


Female 


Mean 


S.D.M 


Diff. 


Mean 


S.D.M 


Diff. 


S.D.diff 


S.D.diff. 


Grade 


+ 21 8 
+ 45 4 
+ 50 4 

+ 36.5 
+ 21.8 

+ 36 5 
+45.4 

+ 36 5 
+ 50 4 


3.40) 
2 92} 
2 05) 

2 281 
3.40/ 

2 28) 
2.92/ 

2 28) 
2 OS/ 


-5.27) 
\ -7 00 
-1.40j 

+3.60 
-2 41 
-4.53 


-86 2 
-84 7 
-74 7 

-85 1 
-86 2 

-85 1 
-84 7 

-85 1 
-74 7 


2 53 
1 47 
1 60 

1.27) 
2.S3/ 

1 271 
1 47/ 

1 27) 
1 60/ 


- 48) 
} -3 82 
-4 62) 

+ .37 
- .19 
-5 10 


High school 


College 


General population 


Grade 


General population 




General population 


College 





It may have been noted that a number of selected adult male 
groups of unquestionably superior education score in scale 
positions unwarranted by the educational trends just noted. In 
the last section of this chapter (the relation of intelligence to 
M-F score) and in later chapters some of the other influences 
to which these divergences must be attributed will be discussed. 
For the present the comparison is between typical, representa- 
tive, unselected populations of adults at the three educational 
levels in question. These comparisons show that for the men 
there are fairly definite trends of score divergence in relation to 
education. 

For the women the differences between the educational groups 
are not as a rule as large as for the men, but they are as persistent 
and tend generally in the same direction. With both sexes a 
higher educational level is associated with greater masculinity 



RELATION OF M-F SCORE TO AGE 145 

in the younger years and throughout life. As we have seen, the 
larger differences for the men are between the grade-school 
groups on the one hand and the high-school and college groups 
on the other; for the women they are between the grade- and 
high-school groups on the one hand and the college groups on the 
other. This sex difference is statistically valid. 

The mean of eighth-grade girls is 10.5 M-F points more feminine 
than that of high-school-junior girls, but the difference is less 
than half its standard error. The difference between the high- 
school juniors and the college sophomores is 24 M-F points, or 
4.1 S.D.flft. The difference between the eighth-grade group and 
college sophomores is 28.1 M-F points, or 6.0 S.D. diffp . As in the 
case of the boys the comparisons are not entirely satisfactory 
because the groups differ in age as well as in education. Certain 
control comparisons may be made between selected groups and 
the unselected general populations (Table 23). For example, 
72 gifted girls in the fourth year of high school whose average 
age is 16 years and 2 months, 3 months younger than the high- 
school-junior girls, score far more masculine than the high school 
juniors. The difference of 27.6 M-F points is four times its S.D. 
On the other hand, 54 delinquent girls whose average age is 
17 years, 4 months are in the ninth school grade and score in 
accordance with this placement between the eighth-grade and 
high-school-junior norms rather than between high-school 
juniors and college sophomores as their age might lead one to 
expect. In contrast to these group placements, which harmonize 
with the expectations based on educational considerations, is the 
mean for a group of 20 part-time schoolgirls whose average age is 
16 years, 5 months and whose educational placement is equiva- 
lent to the sixth or seventh grade but whose mean M-F score 
places them at a point just above the high-school juniors. The 
mean mental age of this group is 11 years, 9J^ months yet they 
score as high as a group of average high-school juniors whose 
mental age is not less than 16 years. 

A group of 28 private-school girls makes an average score of 
65.9 on the M-F scale, approaching the college sophomores and 
differing from the high-school junior rating by 1.5 8.0.^.. 



146 SEX AND PERSONALITY 

These girls are only IS years and 11 months old on the average, 
yet they score far nearer to the gifted girls and the college- 
sophomore norm than to schoolgirls of their own age. Their 
school placement averages at the same point as the high-school 
juniors. In the case of both part-time schoolgirls and the 
private-school girls, neither age nor educational status accounts 
for the M-F score position. The explanation will have to be 
sought in some other factor or factors as yet unrevealed. 

The steps in M-F development at the three educational levels 
show for the decades from the twenties to the eighties the trend 
revealed in the curve for the total population. There is con- 
siderable crossing and recrossing of the lines for the women of 
grade-school and those of high-school training. The grade- 
school women even pass the college norm in the sixties, but in 
general the college norm is for this sex distinct from the other 
two and maintains a distance from each of them in successive 
decades that amounts to 8 or 10 M-F points and ranges up to 
1.3 SJ).^. Combined decade norms emphasize rather than 
lessen these differences and the total divergence of college women 
from high-school women in the populations studied is 10 M-F 
points in the masculine direction, which is 4.6 SD.^. The 
divergence of college women from the population of women with 
a grade-school education is 11.5 M-F points or 3.8 SD.^. A 
general-population group so made up as to match as nearly as 
possible the adult female population of the country as given 
in the United States census differs from this grade-school group 
by only 1 M-F point and from women of high-school education 
by less than half a point, but from the college women's group by 
10.4 M-F points or S.I S-D.^.. 

Summarizing, we may say that mental M-F scores are corre- 
lated with education to an appreciable extent. In the male 
population the order from greatest to least masculinity shows 
first the college group, then the high-school group, and last the 
grade-school group. The differences between these three are 
statistically demonstrable for sufficiently large populations. A 
representative general-population group is more masculine than 
the men of grade-school education, but less masculine than the 



RELATION OF M-F SCORE TO AGE 



147 



men of high-school or college education. In the female popula- 
tions education makes noteworthy differences at all ages for 
which norms have been derived, especially between the more 
feminine groups of grade-school and high-school women on the 
one hand and the more masculine college women on the other. 

TABLE 33. COMPARISON OF MEAN SCORES OF MALES AND FEMALES (a) OF 

GENERAL POPULATIONS AT THREE EDUCATIONAL LEVELS, AND (b) OF A 

COMPOSITE REPRESENTATIVE GENERAL POPULATION 



Populations compared 


Mean 


S.D.M 


Diff. 


Diff. 


S.D.diff. 


General population : 
Adults of grade-school education 
Male 


+ 21.8 
-86.2 

+45 4 

-84 7 

+50 4 

-74.7 


3 40 
2 53 

2 92 
1 47 

2 05 
1 60 


+ 107 9 
+ 130 1 

+ 125 1 


+42 .4 

+39.8 
+48 1 


Female . 


Adults of high-school education 
Male 


Female . 


Adults of college education 
Male . . . 


Female . ... 




A composite representative general-popula- 
tion group (conforming in age and educa- 
tional constitution to the U.S. census 
populations) : 
Male 


+36 5 
-85.1 


2 28 
1 27 


+ 121 6 


+46.6 


Female . 



A representative general population of women scores virtually 
at the same point as do women of grade-school or of high-school 
education, but is significantly more masculine than the women 
of college training. 

Regardless of education, the differences between the mean 
scores of the sexes are very great. The largest difference is 
between the means of the men and women of high-school edu- 
cation, 130 M-F points. The next largest is between the two 
college means, 125 points. The smallest difference, but still a 



148 SEX AND PERSONALITY 

tremendous one, is that between the means for the men and 
women of grade-school education: 108 points. The two typical 
composite sex groups representing as nearly as our data permit 
the general population of the United States differ from one 
another by 122 M-F points. 

THE INFLUENCE OF INTELLIGENCE 

Evidence as to the influence of intelligence upon M-F scores 
will be shown in correlations between M-F scores and scores on 
various intelligence tests, and in comparisons between groups 
otherwise as similar as possible but divergent with respect to 
intelligence. The correlations most useful in this connection 
are those from groups of at least fair size and those having a fairly 
wide range in educational status and mental age. Two such 
correlations have been made, one for each sex for the combined 
eighth-grade and high-school-junior groups. These are between 
M-F scores and mental ages derived from the Stanford Binet, 
the Terman Group Test, and the National Intelligence Test. 
For the boys the correlation is .19 .05; for the girls it is 
.20 .05. The correlations between M-F score and IQ when 
mental age is made constant become, respectively, +.08 .05 
for boys and +.07 .05 for girls, indicating no significant 
relationships. When chronological age is rendered constant, we 
have, respectively, +.15 .05 and +.14 .05, correlations 
which are probably indicative of a small but genuine relationship. 

Correlations for boys and men within small grade or other 
limited groupings give practically no evidence of either positive 
or negative relationship between M-F score and mental age. 
For the girls, however, there is repeated evidence that such a 
relationship is present after the eighth-grade level is passed. 
The coefficient for the high-school juniors is higher than that 
for the combined eighth-grade and high-school-junior groups. 
Student nurses and superior college women show a fairly definite 
tendency for mental level to be correlated positively with M-F 
score. 

A sex difference is probably present in youth in the extent to 
which mental maturity as measured by intelligence tests is 



RELATION OF M-F SCORE TO AGE 149 

TABLE 34. CORRELATIONS BETWEEN M-F SCORES AND INTELLIGENCE SCORES 



Population 


No. 
of 
cases 


C.A. 
range 
in years 


M.A. 
range 
in years 


Correlation 
between M-F 
score and 
mental age 


Males: 
Large group representing a wide range 
Eighth grade and high-school juniors 
Small and select groups, or groups 
representing narrower ranges 
Orphans 


173 

15 
95 
79 

97 
46 
46 

164 

13 
79 
85 

74 
175 


12 to 22 

11 to 16 
12 to 18 
14 to 22 

18 to 32 
19 to 24 
19 to 24 

12 to 19 

12 to 17 
12 to 19 
14 to 19 

18 to 25 
18 to 24 


11 to 21 

11 to 18 
11 to 18 
12 to 21 
Ranges 
of Thorn- 
dike test 
scores 
52 to 111 

59 to 106 

59 to 106 
M.A. 
range 
in years 

11 to 21 

12 to 18 
11 to 17 
13 to 22 
Terman 
group 
score 
90 to 199 


+ .19 .05 

+ .33 .16 
+ .13 .07 
+ 03 07 

+ .00 .07 
+ .20 .10 
+ .24 10 

+ 20 .05 

+ 27 .18 
- 01 .08 
+ .27 .07 

+ 20 .07 
(+ 24 08 
corrected) 
+ 16 05 


Eighth grade 


High-school juniors 

University men (psychology class) 
University men: upper quartile 
scholarship 


University men: lower quartile 
scholarship 

Females : 
Large group representing a wide range 
Eighth grade and high-school juniors 
Small and select groups, or groups 
representing narrower ranges 
Orphans 
Eighth grade . ... 
High-school juniors 


Student nurses 


Superior university women. . . 





associated with M-F score. Coefficients from separate and 
combined groups suggest that for younger boys there would 
probably be found a small but significant correlation between 
M-F score and mental age in a representative population drawn 



150 SEX AND PERSONALITY 

from more than one grade. At the high-school-junior level and 
thereafter it is probable that no such correlation exists. In the 
case of the girls the youngest ages show no significant correlation 
between M-F score and mental age but from the high-school- 
junior level onward and including the college and student-nurse 
populations a positive correlation is found. 

There is nothing in these results to contradict the findings 
respecting the relation of M-F score to chronological age. They 
help to interpret certain puzzling differences between the M-F 
curves of the two sexes. The gain in masculinity by the boys 
from the eighth-grade to the high-school-junior level is not 
duplicated by the girls; it is at this period that mental age is a 
factor in relation to M-F score of boys but not of girls. On the 
other hand, the upward trend of the girls from the third high- 
school year to the second college year is not duplicated by the 
boys; it is at this period that intelligence enters as an important 
factor in the case of girls. Why the intellectual factor should 
exert its influence at different ages in the two sexes is not clear. 

We may next compare the means of groups of corresponding 
ages but of varying intelligence (Tables 22 and 23). The 
average age of the gifted group falls for each sex between sixteen 
and seventeen years, as does also that of the high-school juniors. 
Of the two groups, however, the former is much superior in 
intelligence. The mean M-F score of the gifted boys is +66.2, 
or 5.5 points below the mean of the high-school-junior boys. 
The difference is not statistically significant. However, the 
mean M-F score of the gifted girls is 27.6 points higher than that 
of the high-school-junior girls; the means are, respectively, 
57.3 and 84.8. The difference is four times its S.D. For 
the college students we may compare the M-F scores of those 
who have rated above the average for their institution on the 
Thorndike Intelligence Test for High School Graduates with 
the scores of those who have rated below this average in intelli- 
gence. In the case of men the difference is not significant, the 
high-intelligence group yielding a mean M-F score of +79.5, 
the low-intelligence group a mean of +74. l The corresponding 
means for girls are 24.3 for the high-intelligencq group and 

1 These averages are for the groups A,B,C of Table 18 combined. 



RELATION OF M-F SCORE TO AGE 



151 



46.2 for the low-intelligence group. The difference is 2.6 
times its S.D. 

In the case of our adult groups, since no intelligence test scores 
are available for them, the best we can do is to compare the M-F 
means of the three large populations belonging to three edu- 
cational levels. Owing to the influence of selective elimination 
in the schools, it is well known that men or women who have 



-1-80 



+70 



+60 

Adults 
Col. Ed."* +50 

Adults 
H.S. Ed.~* 

MO 



Adults 
G.S. Ed." 



+30 



+20- 



+ 10 



Superior College Students (l8-25yrs.) 



<- Gifted Boys ( lb| yrs.) - 



-Army Prisoners (26yrs) 
-Preston Delinquents(l7jyrs.) 

Whittier Delinquents (15 yrs) 



Who's Who Men (30-50 yrs) 



Priests in Training (21 yrs.) 



*- Protestant Ministers (20-80 yrs.) 

*" Theological Students (26 yrs.) 
FJG. 7. Age influence upon mean M-F scores of certain male groups. 



graduated from high school excel in average mental ability 
those who have completed only the eighth grade, and that college 
graduates are on the average superior to the general run of high- 
school graduates. The M-F standing of the three adult groups 
is shown in Figs. 7 and 8 in comparison with several other groups. 
We have seen that certain special groups do not follow closely 
the course described by the general population. Comparison 
of the means of these special groups of known characteristics 



152 



SEX AND PERSONALITY 



with the means for the unselected population in part corroborates 
our findings regarding the influence of age, education, and 
intelligence and suggests other factors operative in the M-F 
ratings of the groups in question. Three unusually interesting 
groupings have been selected for each sex. 

The first of these permits of intercomparison of markedly 
superior individuals at three age levels: gifted boys in their teens 



Adults 
Col. Ed." 



-30 



-40 



-50 



-60 



-70 



-80 



Superior College StiHents (20yrs.) 



Who's Who Women{30-50yrs.) 



Gifted G.rls 



Student Nurses < 21 yrs.) 
Adult Nurses (20-bOyrs.) 



Music Students (16-25 yrs) 



Adult Musicians (20-80 yrs.) 



Adults 
H.S. Ed.-> 

Adults -* 
G.S. Ed. -90 
FIG. 8. Age influence upon mean M-F scores of certain female groups. 



whose IQ ratings seven years earlier averaged ISO; superior 
college men in their early twenties whose college aptitude score 
was above the mean for their institution, Who's Who men, an 
unselected group of forty- and fifty-year-old married men listed 
in the well-known volume. The age curve described by the 
means of these three groups is similar to that of the general 
population, but somewhat exaggerated in its main features. 
The gifted boys are less masculine than average schoolboys of 
their age. The superior college students rate highest. The 
Who's Who men are more feminine than the other adult groups 



RELATION OF M-F SCORE TO AGE 153 

which have had college or high-school education and approach 
more nearly than these to the adults of grade-school education. 
Special factors must be operative to make gifted boys and dis- 
tinguished men more feminine, and superior college men more 
masculine, than corresponding age groups of the general 
population. 

A second example is found in three delinquent populations. 
The younger delinquent boys score more feminine than the eighth- 
grade boys who are on the average younger. Older delinquents 
show the expected gain in masculinity from the younger position, 
but they score decidedly more feminine than either high-school 
juniors who are younger or college sophomores who are older. 
Army offenders, in their twenties, are more masculine than either 
of the adolescent groups, but they do not reach the masculinity 
of college sophomores or high-school juniors. 

Three interesting male religious groups are available: a group 
of Catholic priests in training, a group of Protestant theological 
students, and an adult group of Protestant clergy and ministers. 
The Catholic student priests score at a point far less masculine 
than any other male group of their age; in their early twenties 
they are more feminine than the general male population at 
middle life. The Protestant theological students in their middle 
twenties are, however, more feminine than they and exceed in 
femininity the sixty-year-old men of equal education. The adult 
ministerial group is barely more masculine than the Protestant 
theological students and less so than the student priests. They 
exceed in femininity the college men of the seventh decade. It 
is obvious that the age element is practically erased here by other 
influences. Some dominant factors must be present in all three 
groups to make them, without regard to age, conspicuously and 
almost equally lacking in mental masculinity. 

Three groups from the female populations may be briefly 
examined: gifted adolescent girls of average IQ 149; superior 
college women whose intelligence ratings are above the mean for 
their university; and a group of married women in their forties 
and fifties listed in Who's Who. These three groups, like the 
three comparable groups of men, diverge from the general popu- 



154 SEX AND PERSONALITY 

lation norms of corresponding age. The gifted girls are far more 
masculine than the high-school juniors of almost equal age, and 
slightly exceed in masculinity the college women's norm. The 
superior college women exceed the college women's norm in their 
masculinity. They are much nearer in score to the male theo- 
logical students than to the high-school junior girls. The most 
comparable adult group, the Who's Who women, score more 
feminine than they. The Who's Who women are mentally 
more masculine than any of the three main groups of the general 
female population at any age. They are only one-fifth as far 
below the special group of superior college women as Who's Who 
men are below superior college men. Although the gifted boys 
are less masculine than the most comparable group of average 
schoolboys, the gifted girls are more masculine than any group 
of schoolgirls and even exceed average college girls. The superior 
college women exceed the gifted girls in masculinity by more than 
the difference between the superior college men and the gifted 
boys. This comparison indicates again that the factors opera- 
tive in mental masculinity-femininity are not identical for the 
sexes. Intellectual superiority or some factor associated with 
it seems to incline far more to mental masculinity in women than 
in men. In fact, in the Who's Who men intellectual superiority 
strongly favors femininity. 

A second age comparison among the women is between the 
student nurses in the early twenties and the adult group of 
professional female nurses. The student nurses are less masculine 
than any of the college groups of equal age. They are less 
feminine than the (younger) high-school juniors and rate the 
same as older, practicing nurses. The latter score more masculine 
than any of the general population groups of adult women of any 
educational level. Here age is apparently without influence; 
some other factor or factors, perhaps occupational interests and 
habits, are evidently dominant. 

A third age comparison among women may be made between 
music students in their early twenties and a group of adult 
musicians and music teachers. The music students are more 
feminine than the student nurses or any of the groups of college 



RELATION OF M-F SCORE TO AGE 155 

women. They are more masculine than the (younger) high- 
school-junior girls or than any of the decade norms for the general 
population. The musicians and music teachers are as feminine 
as the general female population in middle life. In mental 
masculinity they rate as far below the professional nurses as 
these do below the Who's Who women. In these musical groups 
we find the same age tendency as was noted in the general popu- 
lation, namely an increase of femininity with increase of age. 

Consideration of the above special groups of both sexes shows 
that whereas the age factor is important, other influences are at 
times dominant. Some of the other influences will be treated 
in later chapters. It is sufficient here to point out that numerous 
factors enter to affect the M-F score and that these frequently 
operate dissimilarly for the sexes. 

SUMMARY 

1. The course of mental M-F scores is generally similar for 
the sexes: an initial rise in youth to a more masculine level is 
followed in both sexes by a decline throughout maturity toward 
femininity. 

2. At no time do the means for comparable groups of opposite 
sex approach one another closely. Sex difference measured by 
the M-F score is preserved in all general unselected groups from 
childhood to old age. 

3. The range between the means of the most masculine and 
the least masculine male groups is about twice that between the 
means of the least feminine and the most feminine female groups. 

4. The peak of masculinity in average males is reached in the 
high-school period; the most feminine scores are found in old age. 

5. The greatest extreme of femininity found among females 
is in the eighth grade, the greatest extreme of masculinity in the 
college period. 

6. The feminizing of men in maturity is associated with the 
effect of influences represented in part by a composite of increas- 
ing age and length of married life. For women the relation of 
this composite to the M-F score is less demonstrable. 



156 SEX AND PERSONALITY 

7. In general the same trends are followed and the same sex 
differences shown by populations at three important educational 
levels. 

8. For the men, college and high-school education seem to 
exert a more masculinizing influence than grade-school education. 
For the women, college training apparently has a significantly 
more masculinizing influence than high-school or grade-school 
education. We do not know whether or to what extent selection 
may be responsible for these results. 

9. A general population of males made up in such a way as to 
conform to the United States census proportions for the adult 
population of the country differs by 121.6 M-F points from a 
similarly constituted population of females. This difference is 
approximately twice the range from the most to the least mascu- 
line mean for the general male populations and more than three 
and a half times the range from the most to the least feminine 
mean for the general female populations. 

10. Intelligence is probably positively correlated with mental 
M-F score for both sexes at certain ages. The relationship 
appears to be more pronounced for the youngest boys than for 
the youngest girls, but thereafter more pronounced for the female 
than for the male populations. 

11. Age, education, and intelligence explain only the differ- 
ences between the large general-population groups. Special 
groups selected with respect to occupation or other character- 
istics do not always conform to the expectations based on the 
findings for the general populations. Interesting examples are 
the extremely feminine male religious groups, the Who's Who 
men, so much less masculine than others of equal education, 
and among women the relatively masculine professional female 
nurses and the extremely masculine intellectual groups. 



CHAPTER VIII 

RELATION OF M-F SCORE TO OCCUPATION 

In the preceding chapter we found that age was slightly but 
definitely correlated with M-F score from the youngest to the 
oldest groups. Education and intelligence were also discovered 
to be related factors determining to a measurable extent M-F 
scores of most of the groups studied. The scores of all groups 
were not explainable in the terms of these correlations. It 
appeared that other factors were at work that sometimes entirely 
upset our expectations. For example, two groups of younger men, 
and another averaging at middle age, all well educated, scored 
at approximately the same point as men of the general popula- 
tion at sixty years of age; two groups of women, one in the early 
twenties, the other averaging at middle maturity, scored at the 
same point, which was above all the decade averages for all adult 
women without respect to the level of their education. In both 
these illustrations age was not a common factor. Education was 
probably about equal for the combinations in question, as was 
also intelligence. But as these groups did not score near others 
of the same educational and intelligence level it was manifest 
that some other factor was strongly active. The most natural 
suggestion was that interests and occupations were playing a 
part in influencing the score. One of the nonconforming groups 
was made up of Catholic student priests in their early twenties, 
Protestant theological students in their middle twenties, and 
ministers and clergymen representing all ages from twenty to 
ninety. In spite of the differing ages, sects, and experience of 
these three groups all scored near the same point on the scale, 
and at a position nowhere near the scores for other groups com- 
parable to them in schooling or in intelligence. A similar situa- 
tion was found for two groups of women. Student nurses and 
women engaged in the profession of nursing, although diverse 

157 



158 SEX AND PERSONALITY 

in age, scored at the same point and quite apart from any others 
of comparable intelligence and schooling. In both these com- 
parisons, occupational interest seems a very potent influence. 
The present chapter is devoted to an analysis and discussion of 
M-F scores in the light of the occupations and interests of both 
sexes. 

Total populations for the sexes separately, including all persons 
tested between twenty-five and sixty-five years of age, were 
classified into occupational groups. Several hundred occupations 
were found represented. Many of the groupings were very 
small and wherever possible these were combined into larger 
homogeneous populations. Where there were four or more 
individuals from a given occupation a distribution of the M-F 
scores was made and the mean calculated. The dispersion in 
such groups was much less than is expected in small random 
populations. On the basis of (1) means of the small groups, 
(2) dispersion, and (3) intrinsic homogeneity of activity, sub- 
jectively rated, the small groups were combined and recombined 
into about 45 groups for the men and 74 for the women. These 
were in turn reduced by further amalgamation to 9 principal 
men's groups and 4 most important women's groups. The 
situation regarding occupations for the sexes is quite dissimilar. 
There are many activities for the men any one of which may be 
pursued throughout life with profound interest and devotion. 
For the women there is one chief occupation, that of housewife, 
and there is no other that compares with it in number. Hence, it 
is not possible to discuss occupations for the two sexes in a single 
analysis. We shall present first the relation of men's occupations 
to their standing on the M-F scale and then take up women's 
occupations. 

SCORE TRENDS or MALE OCCUPATIONAL GROUPS 

Each M-F test calls for an entry under the caption, " Profession 
or Occupation." More than 150 different male activities have 
been listed from the information received in answer to this item. 
These occupations have been combined into nine groups by the 
method indicated above. Listed in order from most to least 



RELATION OF M-F SCORE TO OCCUPATION 



159 



TABLE 35. MEAN M-F SCORES OF MALE OCCUPATIONAL GROUPS AND 

SUBGROUPS 



Profession or occupation 


No. 
of 
cases 


Mean 
score 


S.D. 
dist. 


S.D.M 


Stand- 
ard 
score 


A. Professions 
I. Engineers, architects . . 


41 


+ 81 2 


51 9 


8.10 


+ 80 


1. College engineers (no other occupa- 
tion). 
2. College engineers (other occupation) 
3. Architects 


21 
15 

162 


+ 94.3 
+ 70 5 

+ 58 5 
+ 57 3 


45 2 

45 

47 5 


9 87 
11 62 

3 73 


+ 1 05 
+ 60 
+ 40 
+ .35 


1. Bankers 


17 


+ 68 1 


34 3 


8 31 


+ 55 


2. Lawyers 


17 


+ 62.3 


37 


8 97 


+ 45 


3. Realtors .... 
4. Executive managers, superintendents 
5. Salesmen. . . . 
a. Salesmen, special lines . 
b. Salesmen 
c. Traveling salesmen ... 
6 Insurance 


16 
26 
60 
12 

45 

26 


+61 8 
+ 58 9 
+ 58 2 
+63 8 
+ 58 5 
+ 30 5 
+ 40 5 


46 8 
37 7 
50 5 

58 


11.71 
7 39 
6.52 

11 37 


+ .45 
+ 40 
+ 40 
+ 50 
+ 40 
- 10 
+ 05 


7. Manufacturers 




+70 5 






+ 60 


8. Bankers, realtors, insurance, executive 
managers, superintendents. . . . 


85 
166 


+ 55 7 
+45 1 


47 1 
41 6 


5.11 
3 23 


+ 35 
+ 15 


1. Dentists 


14 


+49 1 


50 9 


13 61 


+ 20 


2. Physicians, surgeons 


20 


+45 5 


39 4 


8 82 


+ 15 


3. Teachers 


132 


+ 44.5 


40 8 


3 55 


+ 10 


a. Teachers (college) 
b. Teachers (general) ... . 
c. Teachers (educational admin- 
istration) 
d. Teachers (physical education 
coaches) . 


27 
75 

30 

s 


+ 45 3 
+ 44 6 

+43 8 
+ 82 5 


47 9 
39 4 

36.6 


9.22 
4.55 

6 69 


+ 15 
+ 15 

+ 10 

+ 85 


4 Scientific workers 


9 


+ 54 9 






+ 30 


B. Occupations 


87 


+44 8 


50 1 


5 37 


+ 15 




7 


+64 8 






+ 50 




14 


+ 57 6 






+ 35 




9 


+41 6 






+ 05 


4. Machinists, mechanicians 
5. Chauffeurs, truck drivers 


43 
14 
95 


+40.7 
+ 33 4 
+43.2 


51*3 


5 26 


+ 05 
- 05 
+ 10 


1. Clerks, bookkeepers, office managers, 
secretaries, accountants 
a. Accountants 
b. Clerks 
<;. Bookkeepers . . 
d. Office managers, secretaries . 
2 Postmasters . 


58 
14 
34 

6 
4 


+43 4 
+ 50 5 
+49 3 
+ 30.5 
+ 3.8 
+45 5 


47 2 


6 2 


+ 10 
+ 25 
+ .20 
- 10 
- 60 
+ 15 




9 


+ 43*8 


" * 




+ .10 


4. Merchants, druggists, pharmacists. . . . 
a Druggists pharmacists 


37 
14 


+42 9 
+46 1 


57.0 


9 37 


+ 10 
+ is 


b Mercnants 


23 


+ 37 6 


73 1 


19 6 







4 


+ 10 5 






- 45 




8 


+ 80 






- .50 


VI. Building trades, contractors, painters, 


45 


+ 36 3 


42.1 


6 27 





1 Contractors builders 


14 


+ 51.9 






+ 25 


2. Building trades. . 
3. Painters 
VII Farmers 


23 
8 
58 


+ 32.2 
+ 20.5 
+ 32 9 


43*9 


5 8 


- 05 
- 25 
- 05 


VIII. Policemen, firemen 


23 
10 


+ 279 
+ 28.5 


43 
54 7 


8 97 
17 3 


- 15 
- .10 


2. Policemen ... 


13 


+ 27 4 


31.2 


8.66 


- 15 


C. Vocations 


113 


+ 13 


38.2 


3 59 


- .40 




12 


+ 27.2 


52 2 


15 1 


- 15 




63 


+ 12.7 


35 8 


4 52 


- 40 


3 Artists 


38 


+ 84 


34.9 


5.7 


- 50 















160 



SEX AND PERSONALITY 



masculine, these are as follows: (I) professional engineers and 
architects; (II) professional businessmen: lawyers, salesmen, 
bankers, and executives; (III) professional physical-social 
welfare group: teachers, physicians, surgeons, dentists; (IV) 
mechanical occupations; (V) clerical and mercantile occupations; 



College Athletes 



6en.Pbpi(Coll.Ed)-> 
Gen.PopX30yrs)~* 



Gen.Pbp.<40yrs)- 



Gen.Rop.(50yrs)->. 

Husbands of Who's Who Nbmerr* 

Gen. Pop. (Grade Sch Ed> -> 

fathers of Gifted)^ 



Gen.flop.(60yrs.y> 



+ 90 



:* Engineers and Architects 



+70 



+60 

"* Lawyers, Salesmen, Bankers, 
Executives 

50 Dentists 

,'-Teachers,Physic;ans, Surgeons, 
*- Mechanical Occupations 
"^"Clerks and Merchants 
40 

Building Trades 
<- Farmers 



+30 
Pol 



lice and Firemen 



20 

<- Journalists, Artists, Clergymen 
+ 10 



"10 



Male Inverts 
FIG. 9. Mean M-F scores of occupational and other groups (males). 

(VI) building occupations; (VII) farming occupations; (VIII) 
public-safety occupations: policemen and firemen; (DC) voca- 
tional group: editors, journalists, clergymen, ministers, and 
artists. Reference to Fig. 9 will show the position of these 
groups on the scale with respect to each other and in relation 



RELATION OF M-F SCORE TO OCCUPATION 161 

to norms. There is a tendency for the groups to bunch on the 
scale. For example, farmers and builders are in dose proximity, 
and the three large groups (III) teachers, physicians, surgeons, 
dentists, (IV) mechanical occupations, (V) clerks and merchants 
are within 2.5 M-F points of one another. However, we could 
see no justification on the basis of inherent interest for combining 
any of the nine groups further among themselves. We could 
make a professional, an occupational, and, as we already have 
it, a vocational group, but to do this to any greater extent we 
should have to overlook the occupational differences actually 
indicated. 

Comparing the professions, the occupations, and the vocations 
with one another, we find the three large classes rating in the 
order indicated. The professional rate most masculine and 
somewhat higher than the mean for adults with college training; 
higher also than the mean for age twenty-five of the general 
population. Their position is approximately the same as that of 
the college graduates in their twenties. The occupations rate 
in an intermediate position and at about the same point as the 
general-population group. This places them midway between 
the adults of high-school education and those of grade-school 
training and somewhat below the mean for all adults in their 
forties. They are at the scale position for combined high-school 
and grade-school adults between their forties and fifties. The 
vocations rate lowest. Their position on the scale is far below 
all the general-population averages including that for adults of 
grade-school education. The combined group of editors, 
journalists, clergymen, and artists rates just above the norm for 
sixty-year-old adults. Its mean indicates less masculinity than 
any other general-population group except the sixty-, seventy-, 
and eighty-year-olds. 

Consideration of the positions of the three main activity 
groupings of men shows certain new and important trends. (1) 
Mechanical occupation is strongly masculine in its influence, but 
it does not entirely obliterate the effect of educational level. (2) 
Social or humane pursuits, those concerned with people rather 
than with things, have a feminine influence, but they do not 



162 SEX AND PERSONALITY 

eliminate the effect of education on the score. (3) Culture and 
philanthropy, concerns of the spirit as contrasted with material 
objectives, have a profoundly feminizing influence. Apparently 
this is as strong as the influence of age, and in groups of any size 
it tends to override and obliterate almost completely the effect 
of the educational level. Each of the three trends must be 
considered in greater detail. 

The influence of mechanical occupations is seen in two con- 
spicuous instances in the nine activities. Engineers and archi- 
tects are far more masculine than any other group. The 
difference between them and the professional business group is 
23.9 points or 2.68 S.D. diff .. The difference between them and 
the teachers, physicians, surgeons, and dentists is 4.14 S.D. diff . 
They are 31 points more masculine than the general average of 
men of college education, a difference that is more than 3.5 times 
its S.D. Men in the mechanical occupations are significantly 
less masculine than the engineers and architects'; the difference is 
36.4 points, or 3.74 S.D. diff . But just as the engineers and 
architects score significantly more masculine (1) than either 
of the other professional groupings, (2) than the mean for the 
appropriate decade group, and (3) than the mean for all college- 
trained men, so, too, the mechanical occupations score (a) at the 
masculine limit of the occupational groups, and (6) significantly 
more masculine than the grade-school mean. The respective 
differences among the four occupational groups are neither as 
large nor as significant as those among the (larger) professional 
groups, but the trend from the least masculine groups, firemen 
and policemen, to the most masculine, mechanical occupations, 
seems to support the hypothesis of a mechanical element in 
mental masculinity. The professional and the occupational 
groups meet where the mechanical workers register a high point 
for the occupational groups and where the professional group 
concerned with human welfare (teachers, physicians, surgeons, 
dentists) register the most feminine mean for the professions. 

The second trend noted was for occupations dealing with 
people and social interests to tend in the feminine direction, and 
for those concerned with finance, manufacture, industry, and 



RELATION OF M-F SCORE TO OCCUPATION 163 

mercantile pursuits to tend in the masculine direction. This is 
seen in the placements in the occupational and professional 
groups respectively and in the relation to them of the vocational 
group, especially the clergymen. In both the first-named group- 
ings the interest order is (1) mechanical, (2) business and financial, 
and (3) social and human welfare. The last-named is exemplified 
in the professional groups by the teachers, physicians, surgeons, 
and dentists, and in the occupational groups by the policemen 
and firemen. The most conspicuous example of the human or 
social interest is of course in the clergymen and these rate more 
feminine than any other occupational group except the artists'. 

A third important general trend is the cultural. This is 
indicated, first, by the position of the editors, journalists, clergy- 
men, and artists, and second, by the position of three special 
groups not included in the occupational populations : fathers of 
gifted children, Who's Who men, and husbands of Who's Who 
women. Probably the individuals who make up the three 
last-named groups average in ability and achievement con- 
siderably above any of the occupational groups and probably 
above the other professional groups. The common factor which 
makes Who's Who men, fathers of gifted children, editors, 
journalists, artists, and clergymen score significantly less mascu- 
line than any other group of superior ability cannot be attributed 
to age, education, or intelligence. Subjective considerations 
suggest that it is probably closely related to cultural interests 
and activity. This will also be indicated later by comparison, in 
terms of general populations, of the dominating interest of such 
persons with their M-F scale positions. In the meantime it is 
apparent that the culture factor has for the Who's Who groups, the 
editors, journalists, clergymen, and artists overridden the trends 
for age, education, and intelligence. The principal evidence 
for and against the trends summarized is found in the more 
detailed analysis of each of the nine groups. Table 36 gives the 
statistical intercomparisons of the male occupational groups. 

I. Professional Engineers and Architects. The group of 41 
engineers and architects is made up of (a) 21 engineers, college 
trained and having had no other occupation besides engineering, 



164 



SEX AND PERSONALITY 



S 
o 

I 



i 



I 



I 










-H 

o 



> o I-H w ' 



s 



s 



i O -f 00 

i i 10 ON 



1 1 1 1 1 1 1 1 1 1 1 1 



Sq 

CO 



r co NO OO r>- -^ r*^ !> 



t>- O < 








a-i 
Is 



$ 

ICO 



q 

CO* 



S 
1 

2 

I 





ss 



e . 

fl 









Illllllllllll! 



RELATION OF M-F SCORE TO OCCUPATION 



165 



I 



> 

g 



t 

i-T 

I 



:s 

O '. 
o efl 

* i 


S O 



W 



SI 
8 



S 

a 



i! s 



i .1 '2 - 
- 



a-" 



I! i 



13 

la 



0\ 
+1 
0\ 



-H 
<O 



!9 

Q 



tts 

a 



ttt 
P 



P 
CO 



&' 



!, 

8-S 
3 
g 



1 1 1 1 1 



SO\ **> osi i 
* * ^- ' 






_i *H \O * ^ O 00 



r^ OO -^ rs r-> ^H I-H . 



i i 



\O t^ m ^-4 < 



1 1 1 1 1 1 1 1 






5!^S; 



5 : 

-g tn 



166 



SMX AND PERSONALITY 



a 
i 



t 
i 



- H- 4- + 4- I 



: e > I*: -: ~. ~ v v v v v : v 



I I I 



IT 
II 



I 



I 



I 



1 V V " V V r ** ' 



H-H- + + 



I I I I 



ill 



" r- o ' 

^ ~ & i 




RELATION OF M-F SCORE TO OCCUPATION 167 

(b) 15 engineers college trained and having had other occupations, 
and (c) 5 architects. The subgroups score in the order named. 
The first constituent, the 21 engineers, holds the position of 
highest M-F score among adults. The difference between 
them and the engineers who have had other occupations and 
between both of these and the architects seems to give evidence 
of the strength of the mechanical influence. Architects, being 
akin to artists, should be expected to score less masculine than 
other engineers of equal training, as they actually do. The 5 
manufacturers, who were assigned to the business group, perhaps 
in fact belong to the engineers' group, as their score seems to 
indicate. The proximity should also be noted of men of clear 
mechanical bent who have been assigned to the mechanical 
occupations because they were less well trained than the graduate 
engineers. These include 7 draftsmen and 14 electricians; the 
first are considerably above, the second almost equal to the 
architects in average score. The position of 9 scientific workers, 
at a point a little below the architects, may perhaps also be 
attributed to the mechanical factor. 

The engineers and architects rank well above the professional 
business group; the difference in average score is 23.9 points 
(2.68 S.D.aiff.)- As compared to the seven other professional 
and occupational groups and to the vocational groups, and to the 
principal subgroups, the engineers and architects are significantly 
more masculine, the differences all exceeding the criterion of 
3 S.D.diff.. The difference between this group and the principal 
others is as follows: teachers, physicians, surgeons, and dentists, 
36 points; building trades, 45 points; farmers, 48.3 points; police- 
men and firemen, 53.3 points; and vocational group, 68 points. 
The engineers and architects are undoubtedly high-grade 
individuals with superior training, both general and special, but 
these qualifications do not account for their position with 
respect to the other groups and subgroups. The hypothesis 
of positive correlation between M-F score and mechanical ability 
and interest, a correlation which we shall find persisting at 
every educational level, does, we believe, account for the 
exceptionally masculine score. 



168 SEX AND PERSONALITY 

n. Professional Businessmen. This group is made up of 
17 bankers, 17 lawyers, 16 realtors, 26 executive managers and 
superintendents, 60 salesmen, including traveling salesmen, sales- 
men in special lines, salesmen in general, and 26 insurance 
salesmen and brokers. The six subgroups score on the M-F 
scale from the most to the least masculine in the order named. 
The range is from the bankers at +68.1 to the insurance salesmen 
and brokers at +40.5. It is quite probable that individuals are 
included in some of these groups who have had less than full 
college training. The average is, however, that of college- 
trained individuals and should, we believe, rank the class as a 
professional group. It is somewhat difficult here to identify the 
factors which make for higher and lower scoring. The indi- 
viduals dealing with financial and administrative business on a 
large scale apparently are those who rate more masculine, while 
the businessmen who are concerned with individual contacts 
score more feminine, notably the insurance salesmen. Two 
factors seem to make for greater masculinity: (1) executive and 
administrative work, (2) financial work that is largely free from 
the social aspects. Both of these either demand individuals of 
greater masculinity or else create greater masculinity. Both 
trends are probably included in the element which we have 
called the financial versus the social interest in business and in 
occupations in general. All but one of the subgroups included 
in Class II score well above Class III, the physical-social welfare 
group. In fact, they average at a point approximately halfway 
between the two other professional groups. 

The professional business group is significantly more feminine 
in average (23.9 M-F points) than the engineers and architects. 
The mean difference is 2.68 S.D. difft . All the other large occupa- 
tional groups are more feminine and diverge from Class II as 
follows: mechanical occupations, 12.5 points (1.0 S.D. diff ); 
building trades, 21 points (2.88 SJO.^); farmers, 24.4 points 
(3.55 5.0.^^); policemen and firemen, 29.4 points (3.03 S.D. diff> ); 
and the vocational group, 44.3 points (8.55 S.D. diff J. As far as 
M-F ratings are concerned the superior business group presents 
a marked contrast to the engineers and architects on the one 



RELATION OF M-F SCORE TO OCCUPATION 169 

hand, and to the farmers, policemen and firemen, editors, 
journalists, clergymen, ministers, and artists on the other. It is 
more closely related to the group of physicians, surgeons, dentists, 
and teachers, to the mechanical occupations, to the clerical and 
mercantile group and to the representatives of the building 
trades. The first of the contrasts, that which divides the 
professional businessmen from the engineers and architects, 
seems to involve both a lesser degree of mechanical interest and a 
greater degree of the general social element. The second differ- 
ence, exemplified in the divergence from farmers, policemen, and 
firemen, may well be largely due to the educational factors; the 
professional businessmen are in general more highly trained. As 
contrasted with the farmers the difference seems to be attributable 
to financial versus social-welfare interests. The large divergence 
between the professional businessmen and the vocational groups 
indicates the contrast between financial and cultural interests. 
The superior businessmen rate more masculine than the general 
population, college educated. Their mean is above all the edu- 
cational means and general-population means, and they exceed 
in their masculinity all the general decade norms except that 
for the twenty-year-olds, which they closely approximate. We 
may in summary attribute the position on the M-F scale of 
our professional Class II to superior educational status and to 
financial and administrative versus social or cultural interests. 

III. Professional Physical-social Welfare Class. The group 
of 166 teachers, physicians, surgeons, and dentists is made up of 
three subgroups which rank from most to least masculine as 
follows: 14 dentists, 20 physicians and surgeons, and 132 teachers. 
In view of our previous discussions this arrangement of the sub- 
groups is to be expected. The dentists have doubtless more 
concern with mechanical aspects in their profession than the 
average physician, surgeon, or teacher. The teachers on the 
other hand would probably, in general, have the least interest of 
this kind. The educational difference, if present, would place 
the physicians and surgeons at the highest point, with the 
average for dentists and teachers approximately equal. The 
college professors rate at exactly the same point as the medical 



170 SEX AND PERSONALITY 

men, a combined group of approximately equal college and 
advanced training. We are not inclined to see in the respective 
divergences of the subdivisions of Class III any new contribution 
to our financial-social hypotheses. The pressure of other 
influences than the mechanical adequately accounts for the 
fact that Class III shows no overlap with Class I. The dentists 
who mark the most masculine limit of Class III score at 49.1, 
and the most feminine subgroup of I, the architects, at 58.5, a 
difference of 9.4 score points. 

Class HI rates probably significantly less masculine than 
Class I; the difference in score is 23.9 points (2.68 S.D. <!&.). 
The difference between III and II may also be significant, 
amounting as it does to 12.2 points (2.47 S.D. dlff< ). Educational 
or cultural factors appear to account for the greater femininity of 
Classes IV to IX as compared with the physicians, surgeons, 
dentists, and teachers who make up Class III. From the smallest 
to the largest the differences between III and later groups are as 
follows: IV, mechanical occupations, .3 points (difference less 
than its S.D.); V, clerical and mercantile occupations, 1.8 points 
(difference less than its S.D.) ; VI, building occupations, 8.8 points 
(1.25 S.D. diff> ); VII, farming occupations, 12.2 points (1.85 
S.D. diff< ); VIII, public-safety occupations, 17.2 points (1.80 
S.D. diff> ); and IX, vocational group, 32.1 points (6.65 S.D. dlff> ). 
Class IX, comprising the editors, journalists, clergymen, ministers, 
and artists is the only one of the six that is reliably more femi- 
nine in score than Class III. A complexity of interacting 
factors doubtless accounts for the similarity in score averages 
of the six classes in the central scale positions. As we have 
seen, Class III scores at the extreme feminine position for the 
professions. Class IV, the mechanical occupations, is the most 
masculine within the distinctly occupational groups, and, we 
believe, owes its position primarily to the specific mechanical- 
interest element. The social-welfare interest apparently tends 
about as strongly in the feminine direction as the mechanical 
factor tends in the masculine direction. Each one may be more 
or less counterbalanced by education or some other element. 
The rank of dentists, physicians, surgeons, and teachers with 



RELATION OF M-F SCORE TO OCCUPATION 171 

their superior and specialized training at a point on our scale 
below the norm for men of college education we attribute 
chiefly to the social-welfare factor involved in their occupational 
choices and activities. 

IV. Mechanical Occupations Class. Turning now from the 
professions to the occupations it will be seen that our Class IV, 
mechanical occupations group, is a composite made up of five 
subgroups which rank from most to least masculine as follows: 
7 draftsmen, 14 electricians, 9 engineers (noncollege trained), 43 
machinists and mechanicians, and 14 chauffeurs and truck 
drivers. The relative positions of these subgroups throw further 
light on the strength of the mechanical element which seems 
to have such a positive correlation with M-F score. The more 
technically specialized groups among the five are those whose 
scores rate as most masculine. The educational factor may 
possibly also account in part for the order. The draftsmen 
and the electricians score above the average for Class II, while 
two of the other subgroups score below Class V, clerical and 
mercantile occupations. The low position of chauffeurs and 
truck drivers is thought to be perhaps more a matter of educa- 
tional limitations than of lack in mechanical interest. 

The mechanical occupations rate 36.4 points below the 
engineers and architects, a significant difference attributable to 
(1) general educational difference, and (2) a difference in special- 
ization of mechanical training. They score at a point just below 
the average for men of high-school education. As many of the 
individuals who make up Class IV have completed far less than 
12 school grades, the position of the class on a par with the better 
educated adults once more stresses the masculinity of the 
mechanically minded. Class IV does not diverge significantly 
from the nonmechanical but educationally superior, financial, 
and administrative Class II. It is also practically equal in 
M-F rating to III, the class of teachers, physicians, surgeons, and 
dentists. The feminine social-welfare influence and a higher 
educational status apparently offset the mechanical factor in 
Class IV. The other occupational classes rank successively less 
masculine than IV in the following order and to the extent 



172 SEX AND PERSONALITY 

indicated: V, clerical and mercantile occupations, 1.5 points 
(difference less than its S.D.); VII, farming occupations, 10.4 
points (1 .33 S.D. diff .) ; VIII, public-safety occupations, 15 .4 points 
(1.48 S.D. diff .); and IX, vocational group, 30.3 points (4.76 
S.D. diff .). From this enumeration it appears that IX is the 
only group that is definitely more feminine. However, the 
placement of the mechanical occupations as the most masculine 
among the noncollege-trained male groups when considered with 
other pertinent findings is corroborative of the trends which have 
been pointed out. 

V. Clerical and Mercantile Occupations. A naturally numer- 
ous but well-defined group is the composite V, under clerical and 
mercantile occupations. This is made up of two subclasses, 
(1) clerks, bookkeepers, office managers, secretaries, and account- 
ants, 58 cases; and (2) merchants, druggists, and pharmacists, 
37 cases. From most to least masculine the clerical constituents 
fall in the following order: (1) 14 accountants, 50.5; (2) 34 clerks, 
49.3; (3) 4 bookkeepers, 30.5; and (4) 6 office managers and 
secretaries, 3.8. Ranked from more to less masculine the mer- 
cantile constituents are as follows: (1) 14 druggists and 
pharmacists, 46.1; (2) 23 merchants, 37.6. Mechanical interest 
may be an element in the placement of the accountants, and the 
social interest a factor in the placement of the office managers 
and secretaries. The subgroups are too small for more than 
passing comment. Financial interest is possibly a potent ele- 
ment in the placement of the merchants. Our Class V rates 
just below the mean for the total population. Its position above 
the general population which is made up of persons of grade-school 
and high-school education may be due in part to a somewhat 
better education, in part to financial and administrative interest, 
and in part to a slightly greater element of mechanical interest 
than is present in the three more feminine occupational classes, 
VI, VII, and VIII. However, the differences are too slight to 
bear the weight of much generalization. 

Class V is less masculine than the engineers and architects 
(a significant difference), than professional businessmen (a pos- 
sibly significant difference), than physical-social welfare workers 



RELATION OF M-F SCORE TO OCCUPATION 173 

(difference not significant), and than mechanical occupations 
(difference not significant). On the other hand, the place of 
Class V on our scale shows that it is more masculine than VI, 
building occupations, by 7.0 points (difference less than its S.D.) ; 
than VII, farming occupations, by 10.4 points (1.33 S.D.diff.). 
The vocational group comprised of editors, journalists, clergymen, 
ministers, and artists is dearly and significantly more feminine 
than V. 

In general the placement of Class V may be ascribed in part to 
its level of education, which is just above that of the general 
population, and in part to a balance between financial, adminis- 
trative, and possibly mechanical interest on the one hand against 
the factor of social interest on the other. 

VI. Building Occupations. Three subgroups have been com- 
bined under VI, building occupations. These rank from most 
masculine to least with scores as follows: 14 contractors and 
builders, 51.9; 23 builders, carpenters, etc., 32.2; and 8 painters, 
20.5. Several factors appear to be exemplified in the relative 
ranks of these three subgroups: first, the educational, which 
doubtless in part accounts for the higher placement of the employ- 
ers and contractors; second, the financial and mechanical, each 
of which may well contribute to the differences found; and third, 
the artistic which probably affects the M-F rating of the painters. 
The score differences between the subgroups appear rather large 
but are not significant because of the small size of the groups. 
The trends are similar to those noted among other subgroups and 
add evidence cumulatively to our analysis. 

Class VI rates on the scale approximately at the same position 
as the general population. This is probably a fair placement 
from the educational standpoint, as the individuals making up 
VI are drawn from about the same groupings as our representa- 
tive general population. Within the group itself the masculine 
financial and mechanical factors appear to be balanced by the 
feminine artistic interest represented in the painters. The 
average score for VI is significantly less masculine than that of 
the engineers and architects and probably also than that of the 
professional businessmen. The rating of this group is more 



174 SEX AND PERSONALITY 

feminine (but not significantly so) than the averages for the 
following groups: VII, farming occupation, 3.4 points (less than 
1 S.D. diff> ); VIII, public-safety occupations, 8.4 points (less than 
1 S.D.flff.); IX, vocational group, 23.3 points (3.22 SJO.^.). 
The representatives of the building occupations are thus signifi- 
cantly more masculine than one group only, the vocational. 

VII. Farming Occupations. The group of 58 farmers is proba- 
bly comprised of individuals of quite varying activities, although 
no subgroups could be formed. The individuals here grouped 
are for the most part persons of less than average education. 
Their occupation, if successfully pursued, of course requires a 
moderate facility with tools and some understanding of mechanical 
problems. The chauffeurs and truck drivers rank next above 
them and the carpenters next below. All three probably benefit 
in score about equally from mechanical interest. It is likely 
that this element has pushed all up a little from that position 
natural for men of their average grade of education. Elements 
which may weigh against the mechanical masculinity factor are 
animal husbandry, dairying, and other semi-domestic activities. 

The farmers are significantly less masculine than the engineers 
and architects and than the professional business group. They 
are less masculine, but not significantly so, than the physical- 
social welfare group, 12.2 points (1 .85 S.D. diff J ; than the mechani- 
cal occupations, 11.8 points (1.44 SJ).^ ) ; and than the building 
occupations, 3,4 points (less than 1 S.D.^ff.). Class VII rates 
more masculine than the following: VIII, public-safety occupa- 
tions, 5.0 points (less than 1 S.D.&ff.), and IX, vocational group, 
19.9 points (2.93 S-D.^). Again we find that the occupations 
are not differentiated significantly among one another as far as 
M-F ratings are concerned, but the farmers like the other occupa- 
tional representatives are dearly divergent from the vocational 
group upon which they have traditionally relied for advice and 
direction. 

Vin. Public-safety Occupations. The two subgroups that 
comprise our occupational Class VIII are (1) 10 firemen, average 
score 28.5, and (2) 13 policemen, average score 27.4. The police- 
men and firemen rate at the feminine limit of the occupational 



RELATION OF M-F SCORE TO OCCUPATION 175 

classes. They are probably not less well educated than the 
members of many other occupational groups, and we are inclined 
to interpret the low score as evidence of the social-welfare influ- 
ence which, as already noted, tends to operate in the feminine 
direction. 1 As compared with the educational group to which 
these men belong, their M-F rating is more feminine, but the 
difference is not statistically significant. The average score of the 
group is more feminine than the mean of either of the two very 
masculine professional groups. It is also less masculine, although 
not significantly so, than the mean for physicians, surgeons, 
dentists, teachers and those of the four other occupational classes. 
It is more masculine than the vocational group, but the difference, 
14.9 points, is only 1.54 its standard error. In this fact lies an 
important bit of evidence in favor of our hypothesis regarding 
the effect of the cultural element. It is surely significant that 
in the policemen and firemen the influences (1) of educational 
limitation, (2) of social-welfare interest, and (3) of the absence of 
marked mechanical interest and financial objectives operate to 
produce a score placement which approximates that of the highly 
cultured vocational group. In fact, these civic guardians rate 
practically at the same point as the editors and journalists. 
They rate about halfway between those two superior groups, the 
Who's Who men and the husbands of Who's Who women. It 
thus appears that the positive feminizing influence of culture is 

1 Are policemen really as feminine mentally as this rating indicates, or have we 
through unfortunate chance secured scores from an unusual sampling? In order 
to check the result 15 more policemen, members of the force in a large California 
city, were given the M-F test. Six men failed to complete the test; the 9 who 
finished rated an average score of 4-23.00. The 6 men who completed only part 
of the test showed on the work done approximately the same degree of mas- 
culinity in score as the 9 who finished. The average scores for the public-safety 
group with these 9 men added are as follows: 



Occupational group 


No. of cases 


Mean score 


VIII. Policemen and Firemen: 
1. Firemen 


32 
10 


+26.S 
+28 5 


2. Policemen 


22 


+25 6 









176 SEX AND PERSONALITY 

almost equaled as far as the M-F trait is concerned by a composite 
made up of (1) educational limitation, (2) absence of specific 
mechanical and financial interest, (3) positive social-welfare 
concern, and (4) selective effect of civil-service positions carrying 
low salaries but secure tenure. It would be especially interesting 
to know whether the last-named factor does tend to exclude from 
the ranks of civil-service employees the more ambitious, active, 
masculine type of man. 

IX. Vocational Group. The vocational class is composed of 
three subgroups with number of individuals and average scores 
as follows: (1) 12 editors and journalists, 27.2; (2) 63 clergymen, 
12.7; and (3) 38 artists, 8.4. The interests represented by these 
three classes are most diverse. Placing them in a single voca- 
tional group is, we believe, justified (1) by their score placement 
as most feminine among the occupational classes, (2) by the 
evidence which this placement gives and which is supported 
also by the position of the Who's Who groups that some other 
factor besides education, age, and intelligence almost entirely 
determines their M-F scores. We have called this factor 
"culture." We see it exemplified in the two Who's Who groups, 
men of undoubted superiority and ability, of excellent educational 
training and experience, who, however, rate on the scale at a 
point significantly below the norm for persons of college educa- 
tion. It may well be that not only the positive factor of culture 
but certain negative factors, generally its correlates, are also 
operative in the placement of these most feminine of our normal 
groups. Lack of mechanical, administrative, and financial inter- 
ests are perhaps to be considered here. The social-welfare 
element may be included in the cultural influence or it may be an 
additional factor. 

The vocational group rates at a point near the average score 
for the total population in the sixties. Culture has therefore 
had for this group an effect similar to that of advanced age. 
Class EX is less masculine than any of the eight other professional 
and occupational classes. The differences are as follows: com- 
pared with I, engineers and architects, 68.2 points; II, profes- 
sional businessmen, 44.3 points; III, physical-social welfare 



RELATION OF M-F SCORE TO OCCUPATION 177 

group, 32.1 points; IV, mechanical occupations, 31.8 points; V, 
clerical and mercantile, 30.3 points; VI, building occupations, 
23.3 points; VII, farming occupations, 19.9 points; and 
VIII, policemen and firemen, 14.9 points. The difference is 
significant in every comparison except that with the policemen 
and firemen. Compared with high-school and college levels, 
which best approximate to them in average of educational 
training, they are significantly less masculine. They also differ 
significantly from the general- and from the total-population 
groups. All these comparisons indicate the unique position 
of the vocations on the M-F scale. Whether by reason of nature 
or nurture or by influence of both, the members of Class IX have 
formed mental habits that register on the M-F test in responses 
widely divergent from those of the common run of males. They 
test mentally more feminine than any other male professional or 
occupational group. 1 They are generally recognized as the least 
interested in mechanics, in business, in financial aims, and the 
most concerned with social welfare, religion, and the arts. In 
a word, they represent the "cultured" interests. 

Analysis of the position of the vocational group on the M-F 
scale is not complete without reference to the difference between 
this least masculine of the occupational classes and the M-F aver- 
ages of women's populations. The difference between IX and 
the total population of women of college education is 87.7 M-F 
points. The difference between it and the total population of all 
adult women in our group is 98.1 points. As compared to a 
group of women which rates among the most masculine of all on the 
M-F scale, that is, the 25 women listed in Who's Who, the difference 
is 58.5 points. These figures indicate that the contrast between 
men and women is persistently preserved and that not even the 
strong influence of the culture element which so modifies the 
scores of men in the feminine direction can bring them even 
approximately to the average M-F score of any representative 
group of women. 

1 The single male group which tested more feminine was composed of 82 homo- 
sexuals of the passive type (see Chapter XI). These are not included in the 
occupational classifications. 



178 SEX AND PERSONALITY 

SCORE TRENDS OF FEMALE OCCUPATIONAL GROUPS 

The M-F tests of women between the ages of twenty-five and 
sixty-five make up the groups used for occupational classifica- 
tion. These have been separated into housewives and some 
15 groups of other occupations containing 10 or more persons 
engaged in a particular activity. After a preliminary examina- 
tion of the data each occupational group other than the house- 
wives was subdivided into three sections: single women, married 
women who state that they are actually engaged in the occupa- 
tion or profession in question, and married women who state 
that they are now housewives but that they were formerly 
engaged in this particular occupation. The three sections in 
the subgroups were respectively recombined into three large 
groups for comparison with the occupational, educational, and 
age classes of adult women. For purposes of interoccupational 
comparison the original groups have been used, including the 
persons who state that they are now housewives but were for- 
merly engaged in the activity in question. Means and measures 
of dispersion have been calculated for the 19 occupational groups 
and for various combinations of these. The group of house- 
wives has been subdivided, to add to the separate occupational 
groups in the manner described above. Its menbers have also 
been recombined on the basis of education, into four groups: 
grade school, high school, college for one to three years, and 
college graduates (Table 37). 

Combinations of smaller occupational groups into larger were 
made (1) where the averages were near one another and the dis- 
persions similar, and (2) where homogeneity of activity appeared 
to be present. The combinations on this basis were more difficult 
to make for the women than for the men, as homogeneity of score 
grouping was often not associated with homogeneity of activity. 
For example, hairdressers and dressmakers might seem to be con- 
cerned with equally feminine operations, yet hairdressers are 
among the most masculine, dressmakers among the most feminine 
in M-F rating. Again, one might suppose either that women who 
marry are more feminine than those who do not or else that 



RELATION OF M-F SCORE TO OCCUPATION 



179 



TABLE 37. MEAN M-F SCORES OP WOMEN'S OCCUPATIONAL GROUPS AND 

SUBGROUPS 



Professions and occupations 


No. 
of 
cases 


Mean 
M-F 
score 


S.D.M 


Standard 
score 


I. Professions 


392 


- 74.7 


2 22 


+ 25 


1. High-school and college teachers 
2. Nurses 


14 

64 


- 48.1 
- 64 2 


9 36 

5 27 


+ .85 
+ 45 


3 . Librarians 


20 


- 68 5 


8.88 


+ .35 


4. Teachers (not high school or col- 
lege) 


149 


- 72 1 


3.51 


+ .30 


II. Business occupations 
1. Hairdressers 


303 
10 


- 80 9 
- 69.5 


2 40 
13 57 


+ .05 

+ 35 


2. Secretaries 


43 


- 74 6 


5 74 


+ .20 


3. Secretaries, business administra- 
tors . . 


70 


- 75 2 


4 08 


+ 20 


4. Office managers 


27 


- 76 2 


5 34 


+ 20 


5 Clerks, secretaries . 


105 


- 77.3 


3 84 


+ 15 


6 Bookkeepers, accountants 


46 


- 77 8 


6 49 


+ .15 


7 Bookkeepers, clerks 


108 


- 78 6 


4 04 


+ 15 


8 Clerks typists. 


62 


- 79 1 


5 14 


+ 10 


9. Saleswomen 


22 


- 85 9 


10 60 





10. Stenographers 


70 


- 92 1 


5 44 


- 15 


HI. Housewives. . . 


379 


- 81 6 


1 11 


+ 05 


1. Married women, employed 


179 


- 73 8 


3.30 


+ 25 


2. Housewives, college graduates. . . 
3. Housewives, college education, 
1-3 yrs 


108 
71 


- 75 1 
- 80 1 


3 57 
4 19 


+ .20 
+ .10 


4. Housewives, grade-school educa- 
tion 


10S 


- 82 1 


4 54 


+ .05 


5. Married, formerly employed 
6. Housewives, high-school educa- 
tion ... 


375 
1SS 


- 83.9 
- 85.5 


2 14 
3 29 






7. Housewives, formerly teachers; 
college education (not graduates) 
8. Housewives, formerly elementary 
teachers. 


61 
27 


- 91.1 
-102.8 


5.41 
7 11 


- 10 
- 40 


9. Housewives, formerly domestic 


12 


-107 8 


10 40 


- 50 


Single women employed 


278 


- 74 8 


2 57 


+ 20 


IV Artistic occupations 


W 


- 83 3 


4 71 





1. Artists, decorators, photographers 
2 Musicians music teachers 


14 
41 


- 76.8 
85 6 


10.02 
5 27 


+ .15 

o 


V Domestic occupations 


38 


-100 6 


7 05 


- 35 


1 Dressmakers hairdressers 


31 


- 90 1 


7 03 


10 


2 Dressmakers 


21 


-100.0 


7 21 


- 35 


3. Domestic servants 


17 


-101.3 


13 00 


-.35 



180 



SEX AND PERSONALITY 



marriage has a feminizing effect. But the data show that 
married dressmakers differ from single dressmakers by 10 M-F 
points in the masculine direction, married hairdressers by 
100 M-F points in the same direction, while married office 
managers and business administrators vary in the opposite 
(feminine) direction from single women in the same activities, as 
do also married teachers when compared with single teachers. 
Several trial groupings were made and the resultant means and 
S.D.'s arranged in scale order within the major classifications. 
These comparisons showed new trends quite different from any 
that had been anticipated, yet actually in line with the demon- 
strated influences of age, education, and intelligence. 

TABLE 37a. MEAN M-F SCORES OF ADULTS (WOMEN) BY EDUCATIONAL 

CLASSIFICATION 



Educational group 


No. 
of 
cases 


Mean 
M-F 
score 


S.D.M 


Standard 
score 


A College 


760 


-74 7 


1 60 


+ 20 


B. High school 
C. Grade school 
General population: grade school and 
high school combined 
Total population ... 


820 
287 

1107 
1867 


-84 7 
-86.2 

-85.1 
-80 9 


1 47 
2 53 

1 27 
1 00 






+ 05 













The occupations exhibit the following trends: 

1. The educational-intelligence influence is followed more 
closely by the women's than by the men's groups. The women's 
occupations requiring most training and probably presupposing 
higher intelligence rate most masculine. Conversely, those 
requiring least training and probably including persons of least 
intelligence rate lowest. The two occupational extremes of 
groups of 10 or more persons are held by 14 high-school and 
college teachers, 48.1, and 17 domestic servants, 101.3. The 
difference between these extremes is 53.2 points and in spite of 
the small number of cases in each group it is statistically signifi- 
cant (3.32 S.D.diff.). The educational-intelligence trend is shown 



RELATION OF M-F SCORE TO OCCUPATION 



181 



by the housewives of college education as compared to the grade- 
and high-school housewives. The latter two with marriage have 
exchanged places on the scale. 

2. Leadership, the direction of other people, putting over a 
program, is apparently also a factor in addition to the educational- 
intelligence element. This tendency is seen first in the position 



-10 T 



Superior Athletes 



-20 - 



-30- - 



M.D.orPh.D. 



-40-- 



Whofe Who Women 



-50 - 



-60 - 



Whofe Who Men's Wives 
20 Year Olds 

Gen.Pop. (College Ed.)-'' 
40 Year Olds- 



Gen. Pop. (Grade Sch.) 
60 Year Olds - 
Mothers of Gifted 



Nurses 

Teachefe 

v|-e Professions 

Clerks and Secretaries 
/Business Occupations 
r Housewives 

-Artistic Occupations 



"Sj "*" Dressmakers and Hairdressers 
Stenographers 



-100-T*- Domestic Employees 
FIG. 10. Mean M-F scores of occupational and other groups (female). 

at the top of the women's range of those whose professions and 
occupations involve active dominance of this sort: nurses, libra- 
rians, and teachers in the order named. Similarly the business 
administrators and office managers rate highest among the busi- 
ness and clerical groups, while the position of the stenographers 
is notably low. We are inclined to account mainly on this basis 
for the position of the artists, photographers, and decorators, 



182 SEX AND PERSONALITY 

11 M-F points above the musicians and music teachers. The 
work of the former, if successful, involves essentially more 
administrative activity than that of the latter. The educational- 
intelligence factor may, of course, also be present here. A special 
tendency probably containing many elements related to leader- 
ship is that which we may call the influence of interest in athletic 
sports. This is best typified by a small group of physical 
education teachers who rate at 36.2, which is above the position 
of high-school and college teachers. The element of dominance 
may be the most important here. But since athletic women in 
general, whether captains or not, show the trend we are inclined 
to see in this a trait of the typical independent, self-contained, 
unsentimental woman athlete. 

3. Working outside the home appears to influence the M-F 
score in the masculine direction. 1 Every comparison of large 
groups of housewives with the women in occupations outside the 
home yields a statistically significant difference. Women who 
are not housewives but whose occupation is carried on in their 
own or other people's homes show the same feminine trend. 
For example, the dressmakers and the domestic servants, whose 
horizon is typically circumscribed, are more feminine than the 
hairdressers and cosmeticians, whose occupation is carried on 
away from home in contact with many people and under the 
stimulus of the larger world of business. 

4. In general, marriage has a feminizing influence, or else the 
women who marry are more feminine. That this trend may be 
counterbalanced by others already described is seen in the fact 
that the married women who continue in their previous occu- 
pations score 10 M-F points more masculine than those who give 
up outside activities after marriage. Perhaps, however, it is the 
more able, better trained, and more forceful personalities who 
continue in occupations and professions after assuming home 
duties. It may even be that age alone accounts for this con- 
trast. Comparison of the total population norm with that for 
all housewives and of each educational norm for women adults 

1 But see the discussion on pp. 230jf. 



RELATION OF M-F SCORE TO OCCUPATION 183 

with the corresponding mean for housewives shows, except in the 
case of the grade-school comparison, the housewifely group always 
slightly more feminine. 1 

A corollary to the influence of marriage is a trend exemplified 
primarily by the 78 mothers of gifted children. The members of 
this select group are certainly above the general average in ability 
and training, yet they score more feminine than any of the edu- 
cational or other adult norm groups. It appears that this position 
is due to performance in a single part of the test, word associa- 
tion. The responses indicate an idealistic, romantic attitude, 
especially toward the activities of the home and family. The 
attitude is doubtless a factor in other married groups. It may 
be that it is more pronounced in the better educated and that this, 
in part, accounts for their greater feminization after marriage. 

Summarizing, the trends shown by the female occupational 
groups are these, arranged as nearly as possible in the order of 
their importance: (1) The educational-intelligence influence, 
apparently quite generally operative. (2) The leadership influ- 
ence. Although related to the first trend this appears to have 
additional constituents, thus justifying its naming as a separate 
element. A special form of it is perhaps seen in the athletic- 
response type. (3) The influence of the world of business, and 
interests outside the home. This "broad interest trend" is 
operative in both married and unmarried women who carry on 
activities in business or in the world of ideas. (4) Marital 
influence. This usually operates opposite to the three influences 
enumerated above. It is exemplified in greater degree by 
women of better education than by those of little schooling. A 

1 The exception is perhaps worthy of comment, as it seems to present an 
interesting compensatory trend. In so far as M-F score changes after marriage 
the change can hardly be attributed to the influence of early education or intelli- 
gence. It may be in part associated with age, it is probably to a considerable 
extent connected with interest and occupation. Many of the housewives of 
average or superior education first become interested in household activities 
when they begin to carry them on in their own homes. It is possible that their 
other interests become more circumscribed after marriage. On the other hand, 
the women of grade-school training many of whom have had no occupation 
outside their homes before marriage probably tend to attain a wider mental 
outlook by association and experience in married life. (See also pp. 230jf.) 



184 SEX AND PERSONALITY 

special intensification of it is seen in the response type represent- 
ing an idealization of woman's specific career as wife and mother. 

It is obvious from the above discussion that the trends opera- 
tive in the women's occupational groups are for the most part 
quite unlike those discovered in a study of the men's activities. 
For the men, type of occupational or professional content was 
apparently registered to a considerable degree in M-F score. 
This appears to be less conspicuous in the case of the women, 
whereas intelligence and, broadly speaking, ability are important 
in determining women's scores. A certain analogy may be 
drawn: if mechanical interest is the outstandingly masculine 
trait, domestic activity is the typically feminine one. With the 
men culture operates in a direction opposite to the masculine 
mechanical interest, and with women intelligence, education, and 
breadth of outlook act in the opposite direction to domestic 
interest. Just as the unique position of the engineers implies not 
only a presence of a special interest but also the absence of 
certain cultural interests, so also the position of the most feminine 
of the women's occupational groups implies a limitation in general 
interest and culture as well as a specifically domestic mental 
preoccupation. The similarity between the trends for the sexes 
ends here. For men the influence of intelligence is almost 
negligible, that of education is in the same direction as the typical 
interest (mechanics), while age acts in the reverse direction. For 
the women intelligence is probably an important factor operating 
in the reverse direction from the typical interest (the domestic), 
education is similarly directed, while age tends toward the 
feminine limit. 

Women's professions and occupations may be generally grouped 
under five heads, from least to most feminine in the order named: 
(1) professional, (2) business, (3) housewifely, (4) artistic, and (5) 
gainful domestic. Each of the five groups includes, as stated 
above, women actively engaged in the work, whether single or 
married, and women now housewives who formerly followed the 
occupation. This combining seemed desirable by reason of 
the small size of nearly all the groups. Where groups or com- 
posites were large enough, single and married were originally 



RELATION OF M-F SCORE TO OCCUPATION 185 

isolated for comparison. In Table 3 7 the means for the principal 
subgroups are given. The five composite groups will be con- 
sidered in the order of their rank on the M-F scale from the 
highest to the lowest, with analysis of the subgroup constituents 
of each in turn. 1 

I. The Professions. In the professional group are included 
64 nurses, 20 librarians, and 163 teachers, who rate on the M-F 
scale in the order given. The respective differences between the 
means of these subgroups are all statistically insignificant. The 
subgroups of the three constituents do not follow the same 
order: married nurses, married librarians, and single teachers are 
most masculine; single nurses, housewives formerly librarians, 
and housewives formerly teachers rate most feminine. But as 
married and single women occupationally active outside their 
homes score normally at about the same point, and more mascu- 
line than housewives formerly so engaged, the only divergence 
from the norm is in the position of the unmarried nurses. How- 
ever, the small size of the subgroups in question does not permit 
of any final conclusion being drawn as to the presence or absence 
of real difference trends. 

The nurses' and librarians' groups have not been further 
analyzed. The teacher group has been subdivided wherever a 
sufficient number of cases were available on the basis of (1) level 
of professional activity, (2) education, (3) type of professional 
activity. 

Regarding the first, level of professional activity, the evidence 
is of interest. Among the active teachers a group of 5 single 
high-school and college teachers rates most masculine; next comes 
a group of 9 married high-school teachers, principals, and 
superintendents. These two high-grade groups combined 
approach quite near to the rating of the Who's Who women; the 
scale difference is less than 3 M-F points. Compared with the 
average for all teachers they are definitely more masculine 
(difference 29.2 M-F points, 3.02 S-D.^.). A group of 5 ele- 
mentary teachers rates somewhat lower, at 65.5, but this 
number probably includes only a small fraction of the active 



of the smaller groups not listed in the table are included in the 
discussion. 



186 SEX AND PERSONALITY 

teachers 01 this rank who are included in the general "teacher" 
group. A comparison has been made with the teacher group 
of married women formerly teachers as follows: 13 house- 
wives formerly high-school teachers rate at 80.3, 27 house 
wives formerly elementary teachers rate at 102.8; the difference 
is 22.5 M-F points (1.78 S-D.^.). A combined group of 115 
housewives formerly teachers above the elementary grades rates 
85.3, which is 17.5 M-F points more masculine than the house- 
wives formerly elementary teachers; the difference is 2.12 times 
its S.D. Apparently when the evidence from both the active 
and the married, formerly active teacher groups is taken into 
account, there is considerable basis for belief that higher profes- 
sional status is generally associated with a more masculine score. 

With respect to education the situation is generally similar. 
The occupational classes may be supposed to correspond to suc- 
cessive educational levels. Further data are offered by a com- 
parison of (1) 61 former teachers, college trained but not gradu- 
ates, M-F score 91.1; (2) 23 former teachers, college graduates, 
M-F score 86.0; and (3) 10 teachers whose training extended 
over five years of college or more, M-F score 47.5. While the 
groups are small the steps seem fairly definitely to indicate 
association of M-F rating with educational status. 

Comparison with reference to type of teaching is available to 
a very limited extent. It is presented here purely illustratively 
because of its possible bearing (1) on the special corollary to the 
leadership hypothesis, the athletic-objective-independent type, 
and (2) on the artistic and less dominant, more introverted type. 
A group of 6 physical education teachers rates at the very high 
masculine point of 36.2, while the general average for teachers 
is 77.3 and that for 41 music teachers and musicians is 85.6. 
The type groups are small but they hold similar positions in 
relation to the general population and to the groups of which they 
are a part, as do (1) athletic college women and (2) women music 
students. Hence, it seems that there is a factor which makes 
women physical education teachers rate on the M-F scale at a 
point more masculine than is reached by other women of equal 
training. We have characterized this group as of the athletic- 



RELATION OF M-F SCORE TO OCCUPATION 187 

response type. Similarly, for the music teachers there is some 
element other than education and intelligence which rates 
persons of musical occupation below the general mean. It does 
not appear to be a general cultural factor; culture apparently as 
often tends in the opposite direction, as witness the Who's Who 
group. Subjective analysis of the activities of music teachers 
and the personality types involved leads us to suggest that these 
people represent a fairly superior group of artistically gifted 
persons who tend to lack the dominant and directive, if not 
the independent, character of the other active professional 
women. The music teachers are relatively less interested in 
controlling other people, and relatively more concerned for ideal 
artistic standards. Such a theory would account for the M-F 
rating of this group as compared to the other teachers of equal 
general ability and status. 

The position of the professional group with respect to other 
large groups gives evidence regarding all the points in our 
hypothesis. Comparison with the groups, (1) married and 
working, (2) single and working, (3) married, formerly working, 
and (4) general and total population, shows the professional 
women more masculine than any of the others. This is doubtless 
due to combined factors of age, education, intelligence, activity 
outside the home, and, in a large proportion of the cases, absence 
of marital occupation and domestic interests. Data specific 
to four of the separate heads are as follows: (1) In education and 
intelligence the professional group undoubtedly stands very high. 
Its mean M-F score is above the four other occupational classes, 
although the differences are significant only in the comparisons 
with the housewives (2.78 S.D. diff .) and with the domestic- 
activity group (3.50 S.D.diffJ. The professional mean is at the 
same point as the norm for adult women of college education. 
It is above the means for the other separate and combined edu- 
cational levels; the difference between the professional mean and 
that of women of grade- and high-school training combined is 
statistically significant (4.06 S-D.^.). (2) Leadership and 
dominance are demonstrated by a comparison of the profession- 
ally active part of the group with the average score of persons of 



188 SEX AND PERSONALITY 

approximately equal education, and with the mean for the whole 
group, including those no longer active professionally. The 
active professional women rate 7.5 points above the women of 
college education (2.27 S-D.^) and 7.5 points above the group 
as a whole (2.06 S.D.dj ffe ). (3) The masculine influence from 
working outside the home is found in high scores of the active 
part of the professional group compared with housewives of 
college education. The difference is 7.9 points (1.72 S.D. diff .)- 
If we include the married members no longer active the profes- 
sional mean is barely above that for college housewives. (4) 
Perhaps a feminizing home influence is seen in the lower scores 
of the members who have given up their professions after marriage 
as compared to the higher scores of the active members. 

Our professional group, constituted of elements (1) variously 
professional, (2) more or less well trained, (3) with relatively 
higher or lower intelligence, (4) with more or less well-developed 
qualities of leadership, and (5) more or less influenced by home 
duties, finds placement for its subgroups all the way from the 
highest to the lowest positions on the women's range of the M-F 
scale. As a unit the group scores in the upper fourth of the 
women's range and, when reduced to its most active professional 
element, at a point near the top of all female adult groups. It is 
still far surpassed by the Who's Who women, and rates just less 
masculine than the Who's Who men's wives. Whether married 
or unmarried, the active part of the professional group is made up 
of women capable of independent activity by reason of high men- 
tality, superior training, and successful leadership. These 
traits in so far as they register on the M-F scale have been 
preserved by the married as well as by the single professional 
women in the same way, although in less degree than they have 
been preserved after marriage by the select group of Who's Who 
women. Since the active professional group rates at approxi- 
mately the same point as the wives of Who's Who men we may 
suppose that occupational activity has done for the former, whose 
average intelligence may be somewhat lower, what association, 
native ability, and probably a high type of nonprofessional 
activity have done for the latter. 



RELATION OF M-F SCORE TO OCCUPATION 189 

n. Business Occupations. The business group includes 

(I) 7 auditors, (2) 5 stenographers and secretaries, (3) 10 
hairdressers and cosmeticians, (4) 3 telephone operators, (5) 
4 merchants and proprietors, (6) 43 secretaries, (7) 27 business 
operators and proprietors, (8) 46 bookkeepers and accountants, 
(9) 62 clerks and typists, (10) 22 store clerks and saleswomen, 

(II) 70 stenographers, and (12) 4 bookkeeper-stenographers. 
These rank from most to least masculine in the order 
named. The difference from the highest to the lowest mean 
score is 25 M-F points. The mean for the first five subgroups is 
above the total mean for the professions, the mean for the last 
two is below the mean for the artistic occupations, and the differ- 
ence between these two extremes is probably statistically 
significant. The groups were recombined for Table 37. 

The three subgroups, (a) married and working, (6) single and 
working, and (c) housewives who formerly worked, rate from 
most to least masculine in the order named. In contrast to the 
usual condition the difference between the married and the single 
business women (7.8 M-F points) is almost as large as that 
between the married business women and the married formerly 
in business. This holds in both the lower and the higher testing 
groups of the composite and seems to be generally representative 
of the business class. 

Nine of the 12 business groups score between 70 and 80 on 
the M-F scale, the other three between -85 and 95. The 9 
groups contain all the proprietors and independent operators, and 
the more experienced, better paid, and more responsible types of 
office workers. The M-F range of means within the business 
group is as great as the distance from the mean of the professional 
group to the mean of the domestic-service group. The influences 
of intelligence, education, breadth of interest, and independence of 
activity on M-F position are obvious. The business women are 
6.2 points more feminine than the professional women (1.90 
S-D-diff.)- They are 4.1 points more masculine than the artistic 
group (less than 1 SD.^). They are 6.1 points more feminine 
than the married working women (1.74 S.D.ff.), and 6.1 points 
more feminine than the single working women (1.73 S-D.^.). 



190 SEX AND PERSONALITY 

They rate at exactly the same position as the total test population 
and 4.2 points above the general population (1.54 S-D.^g.). 
The business women test more feminine than college women 
(2.1S S.D.aiff.) but more masculine than high-school (1.35 
S.D. diff .) and grade-school (1.52 S.D.^ff.) groups. Education 
and intelligence thus practically account for the average ratings 
of these women. What some of them have gained in M-F score 
through their more independent activity, others have lost through 
the influence of domestic activity; the result is a score well above 
that of the general population but nearer to it than to the 
professional score. 

DDE. The Housewives. A group of 379 housewives who have 
never had any other occupation includes 105 of grade-school 
education, 155 of high-school education, 11 of high-school and 
normal-school education, and 108 of college education. The 
means for the 4 subgroups rate from most to least masculine 
in the order (a) college, (i) grade school, (c) high school and 
normal, (d) high school. The difference between the first two 
combined and the second two combined is more than 2 S.D. diff .. 
The general tendency found elsewhere for education to correlate 
with M-F score is not eliminated by the element of common 
domestic experience. Comparison with the total population at 
successive educational levels shows that housewives of average 
and above average education are more masculine than the 
corresponding general groups, whereas housewives of grade-school 
education are more feminine. The scale positions of the sub- 
groups of college housewives are in the order (1) college educa- 
tion, five years or more, (2) college graduates (four-year course), 
(3) college one to three years, with a difference of 18.5 points 
between the extremes. 

Compared with the two groups (a) single and working and (b) 
married and working, the housewives and the housewives who 
formerly worked are significantly more feminine. This means 
one of two things: either women whose first and only occupation 
is that of housewife, and women who give up some other occupa- 
tion when they become housewives, are intrinsically more 
feminine than single or married working women, or else they 



RELATION OF M-F SCORE TO OCCUPATION 191 

become more feminine as a result of age plus domesticity. The 
reader may be tempted to say that the mentally feminine woman 
has a mind like that of the typical domestic servant, the mentally 
masculine woman like that of women listed in Who's Who. This 
is true so far as total score on the present M-F test is concerned. 
It would be equally true to say that to be mentally feminine 
means to have a mind like that of typical mothers of intellectually 
gifted children (such mothers averaging high in education and 
intelligence), and to be mentally masculine means to have a mind 
like that of women athletes and women directors of physical 
education. Both statements are true, in terms of the M-F test, 
but either taken alone is misleading. It would be equally true 
to say, in the comparison of male groups, that to be mentally 
feminine means to have a mind like that of the typical male 
homosexual prostitute. We found that so far as total score on 
the M-F test is concerned, the highly feminine score of the latter 
group is closely approximated by Protestant ministers, Protestant 
theological students, and Catholic students training for the 
priesthood. To say that these groups "have minds like" those 
of male homosexual prostitutes would be sufficiently absurd to 
show the danger of generalized characterizations which go beyond 
the specific items of this particular test. 

Only with this emphatic warning do we deem it safe to call the 
reader's attention, as we have in Table 38, to the percentages 
in the housewife and other principal groups who reach the two 
extremes, represented by Who's Who women and domestic 
servants. All the principal trends are exemplified in this table. 
Superior groups have in general the larger percentages of more 
masculine women. The exceptions are the artistic women and 
the mothers of gifted children. The position of the mothers of 
gifted children may reflect the feminizing influence of their 
idealistic attitude toward their privileges as mothers. 

The housewives' average score on the M-F scale is about 
midway within the women's range. The education and probably 
the intelligence of this group is above the general average. But 
domestic occupation plus age counterbalance the masculine 
trend of these factors. Various special populations of housewives 



192 



SEX AND PERSONALITY 



TABLE 38. PER CENT OF WOMEN IN DIFFERENT SUBGROUPS WHO TESTED AT 
THE LEVEL OF Who's Who WOMEN OR OF DOMESTIC SERVANTS 







Per cent more 




Subgroups 


No. 
of 
cases 


masculine than 
the average 
Who's Who 


Per cent more 
feminine than 
the average 
domestic serv- 






women 

(av. -45) 


ant (av. -101) 


Who's Who women 


25 


44 


12 


Wives of Who's Who men 


24 


24 


25 


I. Prof essional group : 


392 


26 


38 


Married and working . . . 


179 


24 


26 


Single and working . . 


278 


26 


27 


II. Business 


303 


20 


31 


III. Housewives 


379 


16 


30 


Housewives who formerly worked 


375 


18 


34 


IV. Artistic occupations: 


55 


14 


29 


Mothers of gifted children 


78 


8 


41 


V. Domestic activities 


17 





55 



illustrate the relative influences of the several masculine and the 
feminine trends in our hypothesis. 1 

IV. Artistic Occupations. The subgroups which form the 
artistic occupation class and rate on the M-F scale from most to 
least masculine, in the order named, are 14 artists, photographers, 
decorators, and 41 musicians and music teachers. The difference 
between the means of the two main subgroups is statistically 
significant. 

The musicians and music teachers, whether single or married, or 
housewives who were formerly engaged in this activity, all rate 
approximately at the same point. The artists, photographers, 
and decorators, on the other hand, show a fairly wide range from 
their most masculine constituent, the married members, to their 
most feminine, the unmarried group. However, all these sub- 
groups are very small and perhaps not truly representative. 
The average of the group as a whole falls below the mean for the 
housewives and between the mean for housewives of grade-school 
training and those of high-school training. The group registers 

1 It should be remembered that the age factor is not constant in the com- 
parisons in Table 38. 



RELATION OF M-F SCORE TO OCCUPATION 193 

at a point intermediate between the total population and the 
general population. It is 8.6 points more feminine than the 
professional average (1.65 S.D.^ff.) and 4.1 points more feminine 
than the business group (less than 1 S.D.^). It is 17.3 points 
more masculine than the domestic-activity group (2.04 S-D.^). 
The deviation from the mean of women of college education is 
the only considerable one in a comparison in terms of education, 
but this is less than 2 SJ).^.. Similar are the differences 
between the artistic group and (1) the women married and work- 
ing and (2) those single and working. We may reasonably 
assume that these women are above the median for intelligence 
and training. Their position below the average on the M-F 
scale may be attributed in part to less than average ability in 
those traits which exhibit themselves in dominating and directing 
other people. Because of the high masculine position of the 
Who's Who women, the factor of culture cannot be thought of as 
explaining the M-F score of the artistic occupations. 

V. Gainful Domestic Activities. This group is composed of 
21 dressmakers and 17 domestic servants whose M-F scores rank 
in the order given. Again, as in the case of the artistic occupa- 
tions, the married members of the group rate most masculine, the 
housewives who were formerly engaged in the occupation next, 
and the single women most feminine. These differences are fairly 
large in the case of the dressmakers but the numbers are too small 
to make the contrast definitely significant. 

The domestic group is more feminine than any of the large 
populations, professional, occupational, educational, and house- 
wifely. The differences are from two to four times their respec- 
tive standard deviations. The very low position of the group 
seems to indicate the nature of the extreme feminine mental type. 
In this instance it signifies less than average intelligence and 
education, and more than average dependence and submission. 
The group stands in direct contrast to the Ph.D.'s and M.D.'s, 
the Who's Who women, women interested in athletic sports, and 
women who are active homosexuals. 

Summarizing: (1) The women's occupations have an M-F 
range about half as great as the men's. (2) The more masculine 



194 SEX AND PERSONALITY 

groups tend to be the better educated and the more intelligent. 
(3) Counter to this tendency operate home interest and house- 
wifely activity. (4) The influence of occupation tends to be 
stronger than the influence of home duties after marriage. 
Women are more apt to score with their occupational group than 
with the housewifely group. (5) Apparently the trend of 
married experience is toward the mean for the total population of 
women : the women belonging to the most masculine occupational 
and educational classes tend to score more feminine if married, 
those who belong to the more feminine occupations tend to score 
more masculine if married. Perhaps this is accounted for by the 
fact that the more masculine scores of the upper group before 
marriage are in part due to the absence of household occupation 
and interest, while the more feminine scores of the lower group 
are due to limitations in general outlook. For the higher testing 
women marriage brings contact with specifically womanly 
activities in the home, in the other case it brings opportunity for 
wider association with less limited ideas. 



CHAPTER IX 

EFFECT OF INTERESTS ON THE M-F SCORE 

SELF-RATED INTERESTS 

Occupation gives a fairly objective expression of interest and 
we have seen that its influence on the M-F score is marked. We 
are now to consider the influence of interests based upon self- 
ratings of " likes " and " dislikes." It is no condemnation of such 
ratings that they are subjective, for they reflect in an important 
way certain temperamental and motivational aspects of the 
personality. When an adult makes such statements as: "I like 
music very much," "I care little or not at all for travel," or "I 
suppose I am less interested in politics than the average," these 
must be taken as significant data; they are probably almost as 
trustworthy as when in response to ordinary questioning an adult 
answers, " I am a carpenter," or " I am a trained nurse." Expres- 
sions of interest no doubt sometimes have a large element of wish 
or fancy and may be quite inaccurate registries of the totality of 
attitudes. But these features by no means brand them as non- 
significant. For example, musicians rate their interest in music 
at the highest point, but so do many persons who are not musi- 
cians. Thus a high rating of interest in music does not necessarily 
connote a musician, but rather an individual who values this type 
of experience; and similarly for other interests. Again tempera- 
mental differences may cause some individuals to rate many 
interests at a high point, others to rate few above average, and 
such deviations have significance for personality. "Likes" and 
"dislikes" represent partial cross-sections of the individual at a 
given time and might well be expected to correlate to some 
extent with the M-F rating, which is also in a sense a cross-section 
measure. 

Several hundred adult subjects rated their interest in each of 

twelve type activities on a five-point scale from the highest point, 

195 



196 SEX AND PERSONALITY 

"having much interest," to the lowest, "having little or none." 1 
The cross-on-a-line method was employed. The twelve activities, 
in the order in which they appeared on the blank, are as follows: 
travel, sports, religion, mechanics, social life, literature, music, 
art, science, politics, domestic arts, and pets. In selecting these 
twelve it was our intent to include some that were typically 
masculine, others typically feminine. 

Correlations were computed between M-F score and interest 
ratings on the five-point scale for several age and educational 
groups of both sexes. These showed for both males and females a 
negative correlation averaging about .40 between masculinity 
and interest in religion. Otherwise the correlations were irregu- 
lar, contradictory, and usually quite low^JThe excessive use 
of the "averape" category in part accounts fftr tfrifl rnndklnm 
An attempt was therefore made to derive more usable results 
by the computation of mean M-F score for subjects rating them- 
selves as having "very much interest" or "little or no interest." 
The discussion that follows is based on the consideration of these 
means. Various comparisons of the means with each other and 
with the averages of unselected high-school and college adult 
populations have been made. The interests found to be char- 
acteristic of relatively higher or lower M-F scores have been 
combined into typical groupings and the relation of these interest 
constellations to the M-F score has been examined. The tend- 
ency to rate many interests as either very strong or very weak 
has been considered with reference to the M-F scores of the 
individuals in the various "enthusiasm" or "indifference" 
groupings. 

THE MASCULINITY-FEMININITY OF MALE INTERESTS 

Two hundred and twelve men 2 of high-school education 
between the ages of twenty-five and sixty-five grouped on the 
basis of their expression of "very much interest" earn the mean 
M-F scores shown in Tables 40a and 406. These are arranged in 

1 The "interests" appeared only on the final edition of the M-F test. Ratings 
are therefore available for only part of our total population. 

1 All of the adult males of high-school education between twenty-five and 
sixty-five who had filled out this particular edition of M-F test blanks. 



EFFECT OF INTERESTS ON THE M-F SCORE 



197 



order from the most masculine to the most feminine. Com- 
parison of this and succeeding rank orders of the interests with 
the original order of presentation for rating (above) on the M-F 
blank shows no evidence that the presentation order influenced 
the ratings. 

TABLE 40a. AVERAGE M-F SCORES OP ADULT MALES OF HIGH-SCHOOL 
EDUCATION EXPRESSING "VERY MUCH INTEREST" IN CERTAIN PURSUITS 



Expressing very 
much interest in 


Mean 
M-F score 


No. of 
cases 


Per cent of 
212 cases 


Standard 
score 1 


1 . Science 


+59.6 


44 


21 


+ .40 


2. Mechanics. 


+54.8 


65 


31 


+ .30 


3. Sports 


+52 7 


73 


34 


+ 30 


4 Travel 


+51 1 


93 


44 


+ 25 


5. Social life 


+46.8 


16 


7.5 


+ .15 


6 Pets 


+46 6 


41 


19 


+ 15 


7. Politics 
8. Literature. . 
9. Music 
10. Religion 
1 1 . Domestic arts 


+46.2 
+42 9 
+38.0 
+31 8 
+28.8 


28 
42 
56 
31 

3 


13 
20 
26 
15 
1 5 


+ .15 
+ .10 
.00 
-.10 
-.10 


12. Art 


+16 2 


21 


10 


-.35 



1 The standard score for the population from which these cases are drawn is +.15. 

TABLE 406. AVERAGE M-F SCORES OF ADULT MALES OF HIGH-SCHOOL 
EDUCATION EXPRESSING "LITTLE OR No INTEREST" IN CERTAIN PURSUITS 



Expressing little 
or no interest in 


Mean 
M-F score 


No. of 
cases 


Per cent of 
212 cases 


Standard 
score 1 


1 . Religion . . . . . ... 


+ 71.6 


57 


27 


+ .65 


2. Domestic arts 
3. Art 


+60 2 

+59 2 


126 
112 


59 
53 


+ .45 
+ 40 


4. Music 


+57 4 


55 


26 


+ .35 


5. Pets . 


+54.9 


77 


36 


+ .30 


6. Literature 


+54 8 


46 


22 


+ .30 


7 Social life 


+53 4 


49 


23 


+ .30 


8 Politics 


+49.2 


63 


30 


+ .20 


9 Science ... 


+47.6 


75 


35 


+ 20 


10 Mechanics 


+40 7 


59 


28 


+ .05 


11 Sports 


+37.4 


26 


12 


.00 


12. Travel 


+19.1 


7 


3 


- 30 













* The standard score for the population from which these cases are drawn is +.15. 



198 SEX AND PERSONALITY 

The individuals in the same population who express "little or 
no interest" in a given one of the 12 activities rate from most to 
least masculine as presented in Table 406. The order in the 
second list may be seen at a glance to be almost the reverse of the 
first. The two rank orders together show (1) the masculine- 
feminine contrast between mechanical or scientific trends on the 
one hand and cultural interests on the other. This result seems 
to reflect the trend already noted in connection with the occupa- 
tions. (2) The tendency for negative interest to rate at a more 
masculine level than positive interest. Item for item the "little 
or no interest" rates regularly higher than the corresponding 
"much interest"; the average difference is 7 M-F points. This 
tendency to express little or no interest is also more popular (is 
shown by more individuals) than positive enthusiasms in this 
group of men. (3) The range from the highest to the lowest 
average is for the positive interests 44 M-F points (.80 standard 
score), for the negative interests 52 M-F points (.95 standard 
score). The smaller of these ranges is more than twice, the 
larger is more than three times, the range from the most to the 
least masculine of the three large occupational groups. 

With respect to the contrast between the scientific-mechanical 
interests and the cultural interests the evidence is nonconflicting. 
The lists agree in denoting as masculine traits (a) much interest in 
.science, mechanics, sports, and travel; and correlatively (6) little 
or no interest in religion, domestic arts, art, music, and literature. 
Relatively feminine traits are similarly found to be (a) much 
interest in art, domestic arts, religion, music, and literature, and 
(b) little or no interest in travel, sports, and mechanics. In 
general this conforms with the popular view. The difference 
between the average M-F score of men who profess very much 
interest in a particular activity and the average of those who 
say they care little or not at all for it is greatest in the case of art; 
then follow in the order named religion, domestic arts, travel, 
music, literature, sports, mechanics, science, pets, social life, and 
politics. For domestic arts the difference is 2.4 SD.^.; for 
religion, art, and travel it is more than 3.6 SD.^. The other 
differences are less than twice their standard deviations. 



EFFECT OF INTERESTS ON THE M-F SCORE 199 

In both interest lists the contrasted trends are definitely 
expressed and in three cases the divergences from the general 
mean for men of comparable educational opportunity are fairly 
demonstrable statistically. These three instances show that men 
who have (1) much interest in art, (2) little interest in religion, 
and (3) little interest in domestic arts are probably definitely 
divergent from the generality; the first in the feminine, the 
second and third in the masculine direction. 

Of the negative interests all except mechanics, sports, an^h 
travel belong to individuals of more than average masculinity < 
It is worthy of note that our results indicate that positive interests 
are not so closely associated with masculinity as the negative 
ones. Positive interest in science is the only item in the first list 
that has a score larger than the lowest of the first six items in the 
second list. This seems to indicate a tendency to opposition or 
outright repudiation of certain interests as a definitely masculine 
trait. In this population it is more masculine not to be interested 
than to be interested, especially along cultural lines, but it should 
be noted that the extreme type of negative interest represented 
by the repudiation of religion and domestic arts, and extreme 
positive interest in art, both connote a demonstrable divergence 
from the generality. 

A further point in this connection is the stronger tendency not 
to be interested than to be interested. Positive are more 
numerous that negative ratings only for medyanirs, sports, and 
travel. Approximately equal numbers are (a) interested, and 
*\b] "not interested, in literature and music. Negative ratings are 
considerably more numerous for science, social life, pets, politics J 
religion, domestic arts, and art. In the case of the last two 
activities the differences are very great (58 and 43 per cent 
respectively). The average number of men who have much 
interest in one or another of the 12 lines is 20 per cent, the average 
number who have little or no interest is 30 per cent. 

Another evidence of the negative relationship between the 
breadth of interest of adult males and their M-F scores is found 
in a comparison of the averages classed by number of interests, 
positive versus negative, without regard to what these interests 



200 



SEX AND PERSONALITY 



are. From Table 41 it is dear, in spite of the fact that the 
differences are not definitely significant statistically, that the more 
numerous the interests the more feminine the score, while the 
repudiation of interests has the opposite effect. The 35 indi- 
viduals who avow no enthusiasms whatever score at +.05 
standard score; those who avow no indifferences score at .25 
standard score. 

TABLE 41. M-F SCORES OF MEN OF HIGH-SCHOOL EDUCATION COMPARED ON 

THE BASIS OF THE NUMBER OF THEIR POSITIVE AND NEGATIVE INTERESTS 



Interest 




InO 


In 1 


In 2 


In 3 


In 4 


In 5 


In 6 


In 7 


direction 




inter- 


or 


or 


or 


or 


or 


or 


or 






ests 


more 


more 


more 


more 


more 


more 


more 


Positive inter- 


N 


35 


177 


143 


97 


59 


29 


12 


6 


ests (having 


Mean 


+ 40 2 


+ 51 6 


+ 51 1 


+ 44 9 


+ 38 6 


+ 30 5 


+ 32 2 




"very much 


S.D.M 


9 94 


3 89 


4 54 


5 75 


7 70 


10 00 


14 62 




interest") 




















Negative in- 




















terests (hav- 


N 


19 


195 


164 


146 


107 


76 


45 


16 


ing "little 


Mean 


+ 21 


+ 52 9 


+ 56 6 


+ 57 4 


+ 53 9 


+ 56 


+ 50 9 


+ 63 


or no inter- 


S.D.M 


13 10 


2 71 


2 96 


4 26 


4 87 


5 78 


7 42 


14 35 


est") 





















The positive-interest averages cover the M-F scale range from 
the vocational mean below to the professional business mean 
above. The negative-interest means extend from about the 
same lower limit to cover more than half of the distance between 
the professional businessmen and the engineers and architects. 
Having little or no interest in religion apparently tends to place 
men of high-school education at the same point on the M-F scale 
as an engineering course places college men. Positive interest in 
art has almost as profound an effect but in the feminine direction. 

The differences between the M-F means of those expressing 
positive and negative interest in each of the twelve pursuits are 
shown in Table 42. The arrangement is by amount of difference 
from the largest positive to the largest negative divergence. In 
the adult male population of high-school graduates those who are 
much interested in travel are definitely more masculine than those 
who have little or no interest in travel. Those who are much 



EFFECT OF INTERESTS ON THE M-F SCORE 



201 



interested in sports, mechanics, and science are possibly also 
more masculine than the men who do not care for these pursuits. 
Politics, social life, and pets are not diagnostic of masculinity or 
femininity in this group. The supporters of these interests rate 
a shade less masculine than the men who repudiate them, but the 
difference is less than its S.D. Interest in literature, music, and 
domestic arts indicates a more feminine trend than lack of interest 
in these matters, while interest in religion or in art is a mark of 
definitely greater femininity than the avowed lack of these inter- 
ests. The largest differences in score are not always the most 
significant, because of the differences in the sizes of the groups. 

TABLE 42. DIFFERENCES BETWEEN AVERAGE M-F SCORES OF ADULT MALES 

OF HIGH-SCHOOL EDUCATION WHO EXPRESS "VERY MUCH INTEREST" 

IN VARIOUS PURSUITS AND THE AVERAGE OF THOSE WHO EXPRESS 

"LITTLE OR No INTEREST" 





M-F score 


Diff. 




diff. 


S.D. dlf 


1. Travel... . 
2. Sports 


+32 1 
+ 15.3 


3.73 
1.44 


3. Mechanics 
4. Science 
5. Politics 
6. Social life 
7. Pets 


+ 14.1 
+ 12.0 
- 3.0 
- 6 6 
- 8 3 


1.53 
1.22 
<1 
<1 
<1 


8. Literature 


-12 


1.10 


9. Music ... 


-19.4 


1.84 


10. Domestic arts. . . . 
1 1 Religion 


-36.4 
-39 8 


2.40 
3 98 


12. Art 


-43 


3 64 



Thus far we have considered only males of high-school edu- 
cation. Do the same or different tendencies appear in other 
populations of adult males? In order to answer this question 
data on interests were collected from a group of men of college 
education by the same method described for the high-school 
adult group. The means for the positive and negative interests 
of this group have been computed and are presented for com- 
parison in Tables 430 and 43J. It should be kept in mind that 
the general college mean has a standard score index of +.25, 



202 



SEX AND PERSONALITY 



TABLE 43a. MEAN M-F SCORES OF ADULT MALES OF COLLEGE EDUCATION 
WHOSE SELF-RATINGS INDICATED "VERY MUCH INTEREST" IN CERTAIN 

PURSUITS 



Expressing very 
much interest in 


Average 
M-F score 


No. of 
cases 


Standard 
score 1 


1. Sports 


68 2 


101 


+ 55 


2. Social life 


66.1 


16 


+ .55 


3. Mechanics ... 


65.1 


92 


+ 50 


4. Science 


64 9 


67 


+ 50 


5. Literature 


61 8 


35 


+ 45 


6. Travel 


61 2 


112 


+ .45 


7. Pets 


56.8 


52 


+ 35 


8. Politics. .. 


51 7 


34 


+ 25 


9. Music .... 


42 1 


68 


+ 10 


10. Art 


41.8 


19 


+ .05 


11. Religion . 


35.2 


30 





12. Domestic arts. 


33 8 


6 


- 05 



1 The standard score of the group from which these cases are drawn is +.25. 

TABLE 43&. MEAN M-F SCORES OF ADULT MALES OF COLLEGE EDUCATION 
WHOSE SELF-RATINGS INDICATED "LITTLE OR No INTEREST" IN CERTAIN 

PURSUITS 



Expressing little 
or no interest in 


Average 
M-F score 


No. of 
cases 


Standard 
score 1 


1. Religion 
2. Art . ... 


76.6 
65 2 


84 
132 


+ 70 
+ 50 


3 . Music 


63 5 


60 


+ 50 


4. Domestic arts . 
5. Social life. . 
6. Politics.... 
7 Pets . 


61.8 
61 6 
60 6 

55 8 


154 

57 
65 
88 


+ 45 
+ 45 
+ 40 

+ 35 


8. Literature. . . 
9. Science 


54 4 
47.7 


54 
85 


+ 30 
+ 20 


10. Mechanics ... 
11. Sports.. 
12. Travel . . 


45 9 

37.8 
28 3 


74 
35 
18 


+ 15 

-.15 



1 The standard score of the group from which these cases are drawn is +.25. 

whereas the general high-school mean index was +.15. An 
upward trend of all the interest means is therefore to be antici- 
pated for the college group as compared with the figures for the 
high-school men. In the high-school group the combined 



EFFECT OF INTERESTS ON THE M-F SCORE 203 

positive-interest means averaged +.10 s.s., which is .15 s.s. 
below the group norm; the combined negative-interest means 
averaged +.24 s.s., which is .09 s.s. above the group norm. In 
the college group the combined positive interests show an s.s. 
index of +.31, or .06 above the group norm; the negative interests 
of +.33, or .08 above the group norm. Positive interests in 
college men are therefore associated with greater masculinity 
than in high-school men, both relatively and absolutely; negative 
interests on the other hand are absolutely but not relatively more 
masculine in college men. Putting it in another way we may say 
that positive interests in college men tend to make them more , 
rather than less masculine than the generality of their group and 
that this is in contrast to the effect in the high-school group. In 
one interest, domestic arts, the opposite effect occurs; in religion, 
music, politics, and science about the expected amount of gain 
occurs. In travel, mechanics, sports, and art the displacement is 
larger than was to be expected; in literature and social life much 
larger. We may say then that almost certainly literature and 
social life have a stronger masculine influence on college than on 
high-school men and that to some extent travel, mechanics, 
sports, and art are associated with greater masculinity in the more 
highly educated group. We may say further that positive 
interests are frequently associated with more than average 
masculinity in men of college training, whereas this is less fre- 
quently the case with men of high-school education. 

As in the population of high-school education so here too the 
active mechanical pursuits hold a masculine position, the cultural 
pursuits a feminine one. Sports, mechanics, and science are 
three of the four most masculine pursuits and repudiation of them 
is a negative trend. Dislike of travel connotes in both popula- 
tions a feminine quality, although great interest in travel is 
for the college men not so masculine a trait, relatively, as it was 
for high-school men. Perhaps the term differs in significance 
for the two groups. In the mind of the man of less education 
"travel" may mean simply going from one place to another, an 
independent exploratory and boyishly active pursuit which might 
perhaps even include hunting and fishing. For the college man 



204 SEX AND PERSONALITY 

it may well have a more intellectual coloring and it is doubtless 
for him as apt to represent pursuit of culture as it is activity on 
its own account. Interest in social life also changes its position 
in the college ratings. It has become one of the two most 
masculine traits. It is perhaps to be expected that a college 
experience will cultivate certain interests and discourage others, 
and that these tendencies would persist in later life. 

It is noteworthy that negative interests as a whole do not 
contrast with positive ones in college men as they did in the men 
of high-school education. The average standard score of high- 
school men with expressed enthusiasms was +.10 s.s., or .05 
below the group norm. The corresponding score for the bar- 
barians of high-school education, the men who had little or no 
interest in the pursuits in question, was +.24 s.s., or .09 above the 
norm. In the college group the enthusiasms rate at +.31 s.s., 
the indifferences at +.33 s.s., both being slightly above the norm 
(+.25) for their group as a whole. We are inclined to see here a 
characteristic difference between men at the two educational 
levels. In the less educated, positive interest tends to be 
feminine, indifference a typical masculine attitude; in the more 
highly educated an attitude of indifference still tends to be 
masculine, but an attitude of positive interest is almost equally 
so. Most masculine of all are still the men who have little or no 
interest in religion, but it is more masculine to have much interest 
in sports, social life, and mechanics than to have little or none in 
any of the other interests. 

Table 44 shows for the college adult group the differences 
between the average scores of the much and the little or not-at-all 
interested for each of the 12 pursuits. The rank order and the 
values are quite similar to those (in Table 42) for the high-school 
adult group. 

The positive-interest means of the college group cover a range 
excluding the vocations but including all the principal professions 
except engineering (Fig. 11). Even the high masculine negative- 
interest means do not reach to the unusual rating of the profes- 
sional engineers. None of the positive or negative interests rates 
nearly so feminine as the vocations of art and the ministry, 



EFFECT OF INTERESTS ON THE M-F SCORE 



205 



TABLE 44. DIFFERENCES BETWEEN AVERAGE M-F SCORES OF ADULT MALES OF 

COLLEGE EDUCATION WHO EXPRESS "VERY MUCH INTEREST" IN 

VARIOUS PURSUITS AND THE AVERAGE OF THOSE WHO EXPRESS "LITTLE 

OR No INTEREST" 



Interests 


M-F score 
diff. 


Diff. 


S.D diff. 


1. Travel. 


+32 9 


2 69 


2. Sports. 


+30 4 


4 06 


3. Mechanics 


+ 19 2 


2 54 


4. Science 


+ 17 2 


2 42 


5. Literature. 


+ 7.4 


<1 


6. Social life 


+ 4 5 


<1 


7. Pets. 


+ 1 


<1 


8. Politics . 


- 8 9 


<1 


9. Domestic arts 


-28 


1 95 


10. Art . 


-23.4 


1 84 


11. Music 


-21 4 


2 73 


12. Religion 


-41 4 


5 28 



although many individuals in the vocational group, especially 
among the clergy, are college men. A point of difference between 
high-school and college men appears in the occupational com- 
parison. Interest, whether positive or negative, is more impor- 
tant than occupation in its relation to the extreme M-F scores of 
men of high-school education. In college men the extreme 
positions, both masculine and feminine, are held by professional 
and not by interest groups. Personal interests are wider than the 
attitudes developed through occupation, but the complex mental 
behavior created through the professions and the vocations may 
develop more conspicuous masculinity or feminity than simple 
individual interests. 

The interests of individuals in such professions as engineer- 
ing, the ministry, or art might, if measured adequately as con- 
stellations instead of separately and incompletely, show a close 
relationship to the M-F score. Since the average scores of indi- 
viduals rated on the basis of single interests show definite 
masculine or feminine trends, the scores of persons who combine 
two or more of the extreme interests might be expected to show 
the same trends intensified. Positive interest in science, 



206 



SEX AND PERSONALITY 



mechanics, sports, or travel is associated with masculinity. Will 
not then great interest in two or more of these indicate still greater 
masculinity? 

Four interest constellations built from the extreme scores in the 
positive- and negative-interest series are analyzed in Table 45. 



Much 
interest 



"75 



+ 70 



+55 



+ 60 



+ 55 



+ 50 



+45 



+40 



+ 35 



+ 30 




Little or no 
interest 

-Religion 



:-Art 
-Music 

"-Social life 
'Domestia 

TOllffCS 

'--Pets 
'-Literature 



\-Science 
-Mechanics 



^--Sports 



\<--Travel 



Arts 



FIG. 11. Mean M-F scores of adult males with much interest or little or no 
interest in various fields. 

Men of high-school education who have (1) positive scientific, 
mechanical, active interests are compared with an intellectually 
related group, namely (2) those who have little or no interest in 
music, religion, domestic arts, and art. (3) Those who have very 
much interest in music, religion, domestic arts, and art are then 
compared with their theoretical correlates, (4) those who have 



EFFECT OF INTERESTS ON THE M-F SCORE 



207 



little or no interest in science, mechanics, sports, or travel. 
The tabular values show that where two interests are combined in 
each of the four groups the negative masculine constellation rates 
highest, the negative feminine constellation second, and the two 
positive groups in their expected order in the last two places. 
The same order is preserved where three interests are combined in 
each constellation, but here the extremes show an intensified 
trend. In the last column the intensification of the extremes 
persists, but here both the positive and the negative masculine 
trends have become increasingly masculine. In the population 
analyzed there is evidence of more or less cumulative effect in each 
of the groupings. It is most pronounced in the negative mascu- 
line group and in the positive feminine one. The difference 
between the average score of 9 men who profess little or no 
interest in music, religion, domestic arts, and art and the average 
score of 11 who profess much interest in three of these interests is 
93 M-F points. The distance is greater than the range from the 
mean of the most masculine to that of the most feminine occu- 
pational, professional, or vocational groups or subgroups. 

TABLE 45. M-F SCORES OF CUMULATIVE INTERESTS IN FOUR TYPICAL 
CONSTELLATIONS (MEN OF HIGH-SCHOOL EDUCATION) 





Combina- 


Combina- 


Combina- 


Interest constellations 


tion of 2 


tion of 3 


tion of 4 




interests 


interests 


interests 


Positive 


Very much interest in 


N 


59 


20 


10 


masculine 


science, mechanics, 


Mean 


+45.4 


+44 5 


+ 68 5 




sports, travel 


S.D.M 


6.92 


10 21 


23 23 


Negative 


Little or no interest in 


N 


74 


43 


9 


masculine 


music, religion, domestic 


Mean 


+60.0 


+57.9 


+ 106 1 




arts, art 


S.D.M 


6.08 


6.96 




Positive 


Very much interest in 


N 


26 


11 


1 


feminine 


music, religion, domestic 


Mean 


+34 4 


+12.3 


- 9.5 




arts, art 


S.D.M 


8 55 


16.35 




Negative 


Little or no interest in 


N 


57 


20 


9 


feminine 


science, mechanics, 


Mean 


+48 4 


+48 5 


+ 63.8 




sports, travel 


S.D.M 


6.90 


10.67 





208 SEX AND PERSONALITY 

Summarizing the results for the adult males, we may say that 
interests as rated by the individuals themselves show a relation- 
ship with the M-F score that agrees with the occupational find- 
ings. Men with objective and mechanical interests tend to rate 
more masculine than the average, men with cultural interests 
including those typical of the vocational group rate more femi- 
nine than the average. Furthermore, the masculine interests 
tend to be typically active, the feminine interests to be typically 
sedentary. The men of college education show the same general 
trend as the adult population with high-school training. The 
various interests hold for them about the same relative place- 
ments, but the entire series is at a more masculine level in line 
with the expected influence of education on the M-F score. 

In the adult high-school population breadth of interest tends to 
feminize, whereas barbarianism, i.e., " having little or no interest," 
even in masculine pursuits, is a masculine trait. The men who 
are much interested in art are probably significantly more femi- 
nine than the population to which they belong, while those who 
have little or no interest in religion or domestic art are also a 
group apart but in the excessively masculine direction. 

A very masculine man in the adult population at the high-school 
level of education is rather inclined not to admit positive enthusi- 
asms. He may have much interest in science, mechanics, sports, 
and perhaps travel, usually has no great interest in social life, 
literature, or pets, and only average interest in politics. He 
has typically little or no interest in religion, domestic arts, art, 
or music. An average man of the same population may have 
much or little interest in politics, he has usually average interest 
in science, mechanics, sports, and travel, and also in religion, 
domestic arts, art, and music. He is prone to like rather than 
dislike social life, literature, and pets. A feminine man of moder- 
ate education is one who is much interested in cultural pursuits, 
art, religion, music, and literature, and who cares little or not at 
all for the active or mechanical interests: travel, sports, mechanics, 
science. He has probably average interest in politics, social life, 
and pets. He does not dislike domestic arts. He has more 
positive than negative interests. 



EFFECT OF INTERESTS ON THE M-F SCORE 



209 



The masculine individual of college education likes sports, 
mechanics, and science. He may dislike social life but he is more 
apt to care for it. He is not likely to be interested in domestic 
arts and he dislikes or is indifferent to religion, art, and music. 
Average interests are for him literature, travel, pets, and politics. 
The well-educated man of average masculinity is not extreme in 
his attitude toward sports, social life, mechanics, science, religion, 
art, music, and domestic arts. He may care much or little for 
literature, pets, and politics. He is more apt than not to be a 
devotee of travel. The feminine college man is a man of culture 
and quiet pursuits. He is not especially attracted to science, 
mechanics, or sports, and may have little interest in travel. He 
neither likes nor dislikes social life, politics, and pets. 

THE MASCULINITY-FEMININITY or FEMALE INTERESTS 

Self-ratings on interests were available from 533 women of 
high-school education ranging in age from 25 to 65. The data 
were obtained in the way described for the men. Averages of the 
M-F scores for the women expressing "very much interest " or 

TABLE 46a. AVERAGE M-F SCORKS OF ADULT FEMALES OF HIGH-SCHOOL 
EDUCATION EXPRESSING "VERY MUCH INTEREST" IN CERTAIN PURSUITS 



Expressing very 
much interest in 


Mean 
M-F score 


No of 
cases 


Per cent of 
533 cases 


Standard 
score 


1 . Mechanics 


-66 3 


19 


3 6 


+ 40 


2. Sports 


-70 4 


104 


19 1 


+ 30 


3. Politics 


-71 


26 


4 9 


-f 30 


4. Pets 


-77 5 


128 


24 


+ 15 


5. Literature 


-79 8 


160 


30 


+ 10 


6. Science 


-82 8 


33 


6 2 


+ 05 


7. Travel 


-85 2 


235 


44 1 





8. Social life 


-86 1 


65 


12 2 


o 


9. Music 


-86 8 


205 


38 4 


o 


10. Domestic arts 


-89 3 


192 


35 9 


- 05 


11 Art 


-90 4 


108 


20 2 


- 10 




-93 1 


139 


26 1 


- 15 













"little or no interest" for each of the 12 type pursuits in the list 
appear in Tables 46a and 46b. As for the men, so too for the 



210 



SEX AND PERSONALITY 



women, the positive-interest series if reversed would read almost 
like the negative series. 

TABLE 466. AVERAGE M-F SCORES OF ADULT FEMALES OP HIGH-SCHOOL 
EDUCATION EXPRESSING "LITTLE OR No INTEREST" IN CERTAIN PURSUITS 



Expressing little 
or no interest in 


Mean 
M-F score 


No. of 
cases 


Per cent of 
533 cases 


Standard 
score 


1. Religion ... 


-60 9 


67 


12.5 


+ .55 


2. Social life 
3. Music 


-74.2 
-84.2 


72 
83 


13 5 
15 5 


+ .25 



4. Politics 


-84 7 


274 


51.4 





5. Mechanics 


-85 3 


376 


70.5 





6. Art 
7. Domestic art 
8. Science ... . 


-85 8 
-87 7 
-88 7 


171 

54 
287 


32 
10 1 

53.8 



-.05 
-.05 


9. Pets 


-91 


178 


28.4 


-.10 


10. Travel... 
11. Sports 
12. Literature 


-91 
-92 2 
-95 9 


50 
161 
44 


9 4 
30.2 
8 3 


-.10 
-.15 
-.25 



Analysis of the two series reveals certain features. (1) The 
masculine-feminine contrast for the women seems to be one 
between active extradomestic interests and the ladylike pursuits 
of home. (2) Positive interests tend on the average to be 
slightly more masculine than negative interests. The difference 
amounts to about 5 M-F points. The negative interests show a 
wider range at both extremes, but their average is at the norm 
for the population to which these women belong. (3) The 
positive-interest range of .55 s.s. extends above the professional 
women's mean, but does not reach the very feminine posi- 
tion of the gainful domestic activities. The negative interest 
range of .80 s.s. extends from above the most masculine pro- 
fession (nursing) to the level of the most feminine occupational 
average. 

The active nondomestic, nonfeminine interests, mechanics, 
sports, and politics connote masculinity in women of moderate 
education. There is no interest whose adherents in this popula- 
tion rate as feminine as the enthusiasts for these male pursuits 
rate in the masculine direction. The three interests that exert a 



EFFECT OF INTERESTS ON THE M-F SCORE 211 

feminine pull are religion, art, and domestic arts. Comparing 
the two lists extended to include even the small differences 
from the general mean, we find the contrasting factors to be wide 
and varied intellectual and scientific interest on the one hand, 
religious, artistic, and domestic pursuits on the other. The 
position of literature definitely apart from religion, art, and music 
is in contrast to the cultural grouping of the feminine interests 
of the men. The negative interests substantiate the interpreta- 
tion. Indifference to religion and social life are definitely 
masculine attitudes. Feminine on the other hand is lack of 
interest in literature, sports, travel, and pets. 

In the women positive interests as such tend to signify mas- 
culinity, negative interests tend to be feminine. There are more 
women in the high-school adult population who express the 
feminine-(negative-) interest trend (28 per cent versus 24). 
There are more who are indifferent to mechanics, sports, politics, 
science, and art than there are adherents of these interests; the 
numbers for pets and social life are about equal; religion, domestic 
arts, music, travel, and literature attract more than they repel. 
The differences are statistically significant (1) between the 
average scores of the high-school adults and the scores of women 
who express much interest in sports; (2) between the average 
scores of the high-school adults and the scores of women who have 
little or no interest in religion. It seems probable that the high 
masculine position of the group interested in mechanics is not due 
to chance and that a larger group would provide a statistically 
significant difference between this average and the general norm. 

There is a certain similarity in the masculinity-femininity of 
specific interests in the two sexes, but whereas in men of moderate 
education positive interest was feminine and negative interest 
masculine, in women of this educational level positive interest is 
masculine, negative interest neutral. Comparison of the effect of 
increasing numbers of positive versus negative interests indicates 
no appreciable differences. Having many or few profound 
interests is clearly a mark of neither a more nor a less feminine 
mental quality in the women of this group. The question may be 
raised, however, whether the neutral values obtained may not be 



212 



SEX AND PERSONALITY 



due to confusion between actual breadth of interest and super- 
ficial enthusiasms. 

TABLE 47. M-F SCORES OP WOMEN OF HIGH-SCHOOL EDUCATION COMPARED 

ON THE BASIS OF THE NUMBER OF THEIR POSITIVE AND NEGATIVE 

INTERESTS 



Interest 




InO 


In 1 


In 2 


In 3 


In 4 


In 5 


In 6 


In 7 


direction 




inter- 


or 


or 


or 


or 


or 


or 


or 






ests 


more 


more 


more 


more 


more 


more 


more 


Positive inter- 


N 


84 


451 


355 


257 


172 


98 


51 


27 


ests (having 


Mean 


-85 4 


-85.3 


-84 7 


-84 


-83.9 


-85 


-88,7 


-91 


"very much 


S.D.M 


4.70 


1.96 


2 24 


2 69 


3.28 


4.61 


5.68 


9 48 


interest") 




















Negative in- 


N 


31 


504 


437 


349 


225 


143 


83 


42 


terests (hav- 


Mean 


-83 7 


-85 2 


-85 2 


-84 9 


-88 


-88 4 


-86 4 


-85.2 


ing "little 


S.D.M 


8.32 


1 87 


2 05 


2 34 


2.78 


3.67 


4 79 


6 44 


or no inter- 




















est") 





















Both the positive- and the negative-interest series cover a 
wider range than the means of the five principal occupational 
groups (Fig. 12). The positive interests extend upward to a point 
usually reached by highly educated and specially trained occupa- 
tional subgroups only. The women of high-school training who 
have much interest in mechanics, sports, or politics are mentally 
considerably more masculine than the professional women as 
a whole. They are far above the mean of business women and 
rate near the office managers, business administrators, and 
secretaries. Only a few small scattered occupational groups are 
more masculine than this, while among the housewives the college 
graduates alone rate higher. The lower end of the range with the 
means of the women much interested in art and religion extends 
to the very feminine score of the stenographers. There is no 
active professional woman's group that rates as far down as this, 
although there are some housewives formerly teachers or 
librarians who score lower. Similarly in the trained occupational 
group it is chiefly the women who because of marriage are no 
longer active occupationally that rate below this point. The 
gainful domestic activities are more feminine, as are also most of 



EFFECT OF INTERESTS ON THE M-F SCORE 



213 



their subgroups, but none of the groups of housewives whose 
activity has never been other than this rates so low. 

The range of the negative interests is even wider. At the mas- 
culine limit the women who have little or no interest in religion 
score above all the professional subgroups; only a few of the sub- 
divisions of these and some small scattering groups are above this 



Much 
Interest 



Little or no 
Interest 




-^-Religion 



t- Domestic Arts 



Social life 



-85 



-90 



*- Music 

Pets 
-Politics 

-Art 

Mechanics 

Sports 

[--Science 



Literature 
Travel 



FIG. 12. Mean M-F scores of adult females with much interest or little or no 
interest in various fields. 

point. Among the trained occupations no more than ten 
individuals reach this level and among the housewives only 
women with five years of college training rate so high. No 
decade level reaches this point. At the opposite limit are the 
women who care little or not at all for literature and these reach 
the very feminine rating of the gainful domestic activities. None 



214 



PERSONALITY 



of the decade norms is as low as this and only a few clerical groups 
or married women formerly teachers or librarians fall lower. 
As with the men, the interest ranges of the women are wider than 
the occupational ranges of the same educational level. Single 
interests are for both sexes in this population as diagnostic of 
M-F score as, or perhaps more so than, occupational classification. 
The differences in score and their significance between the 
means of groups attracted or repelled by a given interest appear 
in Table 48. It is there seen to be definitely more masculine to 
profess interest in sports than to deny such interest. It is 
possibly more masculine, but not certainly so, to have much 
interest in mechanics, literature, politics, or pets. Women who 
care much for travel or for science are perhaps a shade more mas- 
culine than those who do not, but the difference is not statistically 
significant. Little or no interest in domestic arts indicates 
practically no less feminine a woman than much interest. But it 
is true that 36 per cent of the group are among those who have 
great interest in domesticity, while only 10 per cent repudiate this 
interest. Interest in music and in art seems to be characteristic 
of somewhat more feminine women than lack of interest in the 

TABLE 48. DIFFERENCES BETWEEN AVERAGE M-F SCORES OF ADULT FEMALES 

OF HIGH-SCHOOL EDUCATION WHO EXPRESS "VERY MUCH INTEREST" 

IN VARIOUS PURSUITS AND THE AVERAGE OF THOSE WHO HAVE 

"LITTLE OR No INTEREST" 



Interests 


M-F score 


Diff. 




diff. 


S.D. diff . 


1. Sports 


+21.9 


4 18 


2. Mechanics 


+ 19.0 


1 76 


3 . Literature 


+ 16.1 


2.44 


4. Politics 


+ 13.6 


1.49 


5. Pets 


+ 13.5 


2.81 


6. Travel 


+ 5.8 


<1 


7. Science ... 


+ 58 


<1 


8. Domestic arts 


- 1.6 


<1 


9. Music . . 


- 2.6 


<1 


10. Art 


- 4 6 


<1 


11. Social life 


-11 9 


1.74 


12 Religion 


-32 3 


4.59 









EFFECT OF INTERESTS ON THE M-F SCORE 



215 



same, but the difference in the present group is unreliable. 
Interest in social life is possibly more indicative of femininity 
than its opposite, while interest in religion is typically the extreme 
feminine interest, its opposite being associated with a definitely 
more masculine average score. Again the small size of the 
groups makes most of the differences unreliable statistically. 

We turn next to the corresponding comparisons for women with 
one or more years of college training. The means for the two 
interest series appear in Tables 49a and 496, where they may be 
compared with the general college women's group norm repre- 
sented by the standard score +.25. 

TABLE 49a. AVERAGE M-F SCORES OF ADULT FEMALES OF COLLEGE EDUCATION 
WHOSE SELF-RATINGS INDICATE "VERY MUCH INTEREST" IN CERTAIN PURSUITS 



Expressing very 
much interest in 


Average 
M-F score 


No. of 
cases 


Per cent of 
533 cases 


Standard 
score 1 


1. Science 


-48 3 


42 


7 9 


+ 85 


2. Mechanics 


-54 5 


10 


1 9 


+ .70 


3. Sports 


-58.5 


72 


13.5 


+ 60 


4. Politics... 


-59.5 


30 


5.6 


+ 60 


5. Pets 


-68 3 


58 


10 9 


+ .35 


6. Travel. 


-70 4 


207 


38 8 


+ .30 


7. Social life 


-70.5 


62 


11.6 


+ .30 


8. Literature 


-71 


169 


31 7 


+ .30 


9. Art. 


-72 1 


84 


15 8 


+ .30 


10. Music. 


-72 3 


123 


23 1 


+ .30 


11. Domestic arts 


-77.0 


96 


18.0 


+ .15 


12. Religion.. 


-87.3 


87 


16.3 


-.05 



iThe average M-F score for the college women's population is +.25. 

The positive interests show gains over and above those to be 
expected from the masculine influence of higher education. The 
entire series averages a standard score of +.38, or .13 higher 
than the norm, as compared to +.05 for the moderately educated 
women. Interest in science has advanced to the most masculine 
position of all. The position of literature has become more 
feminine. Otherwise the relative standings resemble those of the 
high-school population. The order still represents the tradi- 
tional expectation: women with scientific, active, nondomestic 



216 



SEX AND PERSONALITY 



interests are masculine in comparison with the feminine adherents 
of the fine arts, literature, and religion. 

TABLE 496. AVERAGE M-F SCORES OF ADULT FEMALES OF COLLEGE EDUCATION 
WHOSE SELF-RATINGS INDICATE "LITTLE OR No INTEREST" IN CERTAIN 

PURSUITS 



Expressing little 
or no interest in 


Average 
M-F score 


No. of 
cases 


Per cent of 
533 cases 


Standard 
score 1 


1 Religion 


-44 3 


61 


11.4 


+ 95 


2. Domestic arts . ... 


-53 9 


46 


8.6 


+ .70 


3 Social life . . 


-58 3 


34 


6 6 


+ 60 


4 Music 


-69 4 


31 


5.8 


+ 35 


5. Pets 
6 Politics 


-72.1 
-73 7 


121 
146 


22.7 
27 4 


+ .30 
+ 25 


7. Art . ... 


-76.2 


59 


11.1 


+ .20 


8. Mechanics 
9. Sports ... 
10. Science 


-78 1 
-79 6 
-81 9 


228 
77 
123 


42.8 
14.4 
23.1 


+ 15 
+ 10 
+ .05 


11. Literature 
12. Travel 


-88 1 
-89 3 


11 

23 


2 1 
4.3 


- 05 
-.05 













1 The average M-F score for the college women's population is +.25. 

The negative interests are also if anything masculine rather 
than feminine in their influence as a whole. The range is slightly 
greater than that of the positive interests. Again the mas- 
culinity of indifference to social life is evident, while indifference 
to travel appears as a feminine indicator. 

The ranges of both interest series include all the professional 
means, and of the occupational means all except that of the gain- 
ful domestic activities (see Fig. 12). Interests, whether positive 
or negative, have little tendency to feminize the minds of these 
highly educated women to the degree exemplified by the dress- 
makers and domestics. Inasmuch as a large proportion of the 
college group is married, this fact seems important. Most femi- 
nine of all at the college level are the women without interest 
in four active or intellectual pursuits: travel, literature, science, 
and sports. 

Masculine college men care for sports, social life, mechanics, 
and science; masculine college women for science, mechanics, 
sports, and politics. The difference is in the more intellectual 



EFFECT OF INTERESTS ON THE M-F SCORE 



217 



fourth interest of the women. The masculine negative interests 
bring out the same contrast. The masculine college man is 
active and social, the masculine college woman is intellectual and 
nonsocial. The feminine positive interests are alike for the 
sexes. The feminine negative interests are found to be, for the 
men, indifference to mechanics, for the women indifference to 
literature. 

Both positive and negative interests tend to be associated with 
more than average masculinity in college women. Absence of 
interest in social life is correlated with masculinity in women at 
two educational levels, while interests indicative of intellectual 
breadth are also associated with the masculine trend. The latter 
tendency is more pronounced in the more highly educated women. 
The differences between the positive- and the negative-interest 
means are mainly inconclusive, statistically (Table SO). The 
trends that they suggest further illustrate the characteristics 
found in our analysis. 

TABLE 50. DIFFERENCES BETWEEN AVERAGE M-F SCORES OF ADULT FEMALES 

OF COLLEGE EDUCATION WHO EXPRESS "VERY MUCH INTEREST" IN 

VARIOUS PURSUITS AND THE AVERAGE OF THOSE WHO HAVE "LITTLE 

OR No INTEREST" 





M-F score 


Diff. 




Diff. 


S.D.drff. 


1. Science. . . 


+ 33 6 


3.83 


2. Mechanics. ... 


+23 6 


2.18 


3. Sports. .... 


+21 


2 73 


4. Travel 


+ 18 8 


2 36 


5. Literature 


+ 17 2 


<1 


6. Politics.. . . 


+ 14 2 


1.77 


7. Art 


+ 4.1 


<1 


8. Pets 


+ 3.8 


<1 


9. Music 


- 2.9 


<1 


10. Social life 


-12.2 


1.28 


11 Domestic arts 


-23.2 


2.91 


12. Religion 


-43 


4.27 









There are individual women who score more masculine than the 
mean for the most masculine profession, higher indeed than the 
most masculine interest whether positive or negative. There are 



218 



SEX AND PERSONALITY 



others who score more feminine than the most feminine occupa- 
tion, and beyond the most feminine of the interest ratings. M-F 
scores of the women who combine interest in two or more of the 
distinctively masculine interests (a) positive, and (6) negative, 
and scores of those who combine interest in two or more of the 
characteristically feminine interests (c) positive and (d) negative, 
help to indicate the sorts of persons who score in extreme positions. 

TABLE 51. M-F SCORES OF CUMULATIVE INTERESTS IN FOUR TYPICAL 
CONSTELLATIONS (WOMEN OF HIGH-SCHOOL EDUCATION) 





Combina- 


Combina- 


Combina- 


Interest constellations 


tion of 2 


tion of 3 


tion of 4 




interests 


interests 


interests 


Positive 


Very much interest in 


N 


55 


5 





masculine 


mechanics, sports, poli- 


Mean 


-52.1 


-73 5 






tics, pets 


S.D.M 


5 68 






Negative 


Little or no interest in 


N 


75 


21 


3 


masculine 


religion, social life, 


Mean 


-78 3 


-79 


- 62 8 




music, art 


S.D.M 


6 04 


8 69 




Positive 


Very much interest in 


N 


124 


61 


15 


feminine 


religion, art, domestic 


Mean 


-88 9 


-89 2 


-121 5 




arts, music 


S.D.M 


3 31 


5.07 


15 18 


Negative 


Little or no interest in 


N 


90 


18 


6 


feminine 


literature, sports, travel, 


Mean 


-94 2 


-88 4 


-102 8 




pets 


S.D.M 


3 86 


9 87 





The masculine constellations show consistently more masculine 
scores than the feminine groupings. The positive-interest 
combinations are the most masculine of all. The two feminine 
combination series persist in the femininity of their scores from 
the two- to the four-interest groupings. Four contradictions are 
suggested which the small numbers are insufficient to validate: 
(1) it is less masculine to have 3 than 2 positive masculine 
interests and (2) it is less masculine to have 4 than 2 or 3 negative 
masculine interests; (3) it is more feminine to have 4 positive 
feminine interests than to have 2 or 3 ; and (4) it is less feminine 
to have 4 negative feminine interests than to have 4 positive 



EFFECT OF INTERESTS ON THE M-F SCORE 219 

feminine interests. Although single positive interests are more 
often masculine than feminine in women of high-school education, 
there is at least a suggestion that in cumulation they may be 
feminine in influence. In other words, a combination in one 
person of several positive interests, each of which may have 
by itself a positive M-F value, connotes in the combination a 
feminine trend. Perhaps this occurs because these women are 
more enthusiastic than interested. The converse is also sug- 
gested: women whose single negative interests are associated 
with a certain score show less increase in the expected direction 
with combined-interest scorings. The explanation is offered 
tentatively that while negative interests represent singly a femi- 
nine trend, grouped they may incline to the masculine trait of 
opposition and indifference found to be rather typical of the 
average man in this population. 

The relation between M-F score and women's self-ratings as to 
interests may be summed up by a few brief characterizations : The 
mentally more masculine woman of high-school education tends to 
care intensely for few rather than many pursuits. She is usually 
interested in sports or politics, and perhaps also in pets. She also 
tends to be intense in her dislikes. She is apt to have little or no 
interest in religion or in social life. As to domestic arts, music, 
travel, science, and literature, her interests are average. The 
mentally most feminine woman tends to be indifferent to few 
interests; on the other hand she tends to many enthusiasms. She 
is interested in religion, but not in literature or sports; her interest 
in social life is average or above. Otherwise her interests are not 
specific. The average woman of the group expresses moderate 
interest in art, domestic arts, music, travel, and science. Her 
interest in religion and in sports is average, while literature, social 
life, and to a less extent travel and pets are fairly sure to attract 
her. She is unlikely to be intense in her interests or to limit their 
number too carefully whether positive or negative. She is apt 
rather to express a number of varied enthusiasms. 

The woman of college education whose score is masculine tends 
to positive as often as or oftener than to negative interests. She 
cares for science, mechanics, sports, politics, or pets; she is indiffer- 



220 SEX AND PERSONALITY 

ent to religion, domestic arts, and social life. She does not dislike 
travel, literature, or art. The college woman at the average of 
masculinity and femininity tends to have some positive and some 
negative interests: she may like or dislike music, art, and pets; 
she may care much for travel or social life, she is not apt to dislike 
either; she may care little for politics, she is not apt to be greatly 
attracted by it. She does not care much for nor is she repelled 
by science, mechanics, sports, religion, or domestic arts. The 
feminine college woman is religious and domestic, especially 
religious. She does not usually care for travel, literature, 
science, sports, or mechanics. Her interest in social life is posi- 
tive rather than negative. She dislikes rather than likes politics; 
she has perhaps average interest in pets. 

The high-school and the college level both show that mental 
masculinity in women is associated (1) with the profession of 
positive interests usually regarded as typically male, (2) with the 
assertion of indifference to the acceptedly feminine arts and pur- 
suits. They show that mental femininity is associated in women 
with pursuit of the feminine and domestic interests and denial 
of interest in the active and typically masculine pursuits. In 
both groups of women positive interests tend to be more mascu- 
line than negative ones. The difference between the higher and 
the lower educational levels consists in the greater intellectuality 
associated with the masculine interests of the one, as compared 
to the greater activity in the masculine interests of the other. 
The meaning of the M-F score is further defined by these 
comparisons. 

INTERESTS or MEN AND WOMEN COMPARED 

The interest most popular among men is not the most mas- 
culine interest, the least popular is not the most feminine. The 
interest which most often repels men is not the most feminine, nor 
is that which least often repels them the most masculine. 
Between popularity of interest and M-F score there appears to be 
no high correlation for either sex. The percentages are shown 
graphically for the adults of high-school education of both sexes 
in Fig. 13. Travel attracts the largest percentage of both 



EFFECT OF INTERESTS ON THE M-F SCORE 



221 



sexes, yet it is only the fourth most masculine interest for the men 
and holds the seventh place for the women. Art, the most 
feminine male interest, attracts ,10 per cent of the men, thus 
rating as more popular than social life (which is most masculine) 
and more popular than the only less feminine domestic arts. 
Relative as well as absolute popularity of interests appears to be 

10 20 30 40 50 60 70 80 90 100 

Much interest-*- -^Little or no interest 
100 90 80 70 60 50 40 30 20 10 









\ 








U/i 








Science 


' 






j 














Mechanics 


_j 






/^ 










9 


^?^7 


Sports 


1 


_, 

r^ 
















V 


Travel 


Wjj>X 








d 










C 


Social life 


1 


\ 














*1 


1 


Pets 


',/VxVy 


L 


3 
















Politics 


W 




















Literature 






~i 














1 


Music 


r///Sj 


L 














^ 


1 


Religion 


\ 
















t, 


1 


Domestic arts 


*SS72. 








t// 












Art 














^ t 









LEGEND 
^Z3 Men 
' i Women 

FIG. 13. Percentages of men and women professing much interest or little or no 
interest in various fields. 

a factor conditioning the masculinity-femininity of the interest. 
The ratio of the popularity in one sex to that in the other accounts 
for placements where simple numbers are not sufficient to explain 
the rank order. 

Masculinity in men at either of two educational levels is asso- 
ciated with positive interest in active and mechanical pursuits; 
it is also associated with indifference to artistic and cultural 
pursuits. The trends are accentuated in both directions in the 
less educated population where also the barbarian response, 



222 SEX AND PERSONALITY 

indifference as such, is strongly masculine. Femininity in men is 
the reverse of masculinity in its relationship to interests. It is 
marked by the presence of positive interests as such, the cultural 
emphasis, and usually by the absence of the more active and 
mechanical pursuits. 

Masculinity in women, whether of high-school or college edu- 
cation, is associated with positive rather than negative interests, 
with activity, and, in the more highly educated especially, with 
intellectuality. Femininity in women shows the reverse of the 
masculine trends. It appears associated with less intense posi- 
tive interests, and these are directed toward the arts of home and 
social life. It may be found also in women who profess a large 
number of enthusiasms, for even those interests which singly 
are masculine prove to be feminine when clustered. Mental 
femininity is expressed also in indifference to the active, scientific, 
and, in the better educated, intellectual interests. The mascu- 
line man has typically masculine interests. The feminine man 
has cultural interests. The feminine woman has typically 
feminine interests. The masculine woman has intellectual 
interests. 



CHAPTER X 

RELATION OF M-F SCORE TO THE DOMESTIC MILIEU 

Youth is masculine, age feminine. Within the area repre- 
sented by the M-F test, culture tends to make men's minds 
resemble women's, intelligence and education to make women's 
minds resemble men's. But every feminine man's score does not 
come from a cultured greybeard, and masculine men's scores do 
not necessarily indicate hard-boiled philistinism; nor are all 
feminine-scoring women dull and uneducated. Some elements 
of environment and constitution which have been found related 
positively or negatively with the total M-F score are discussed in 
this chapter, including the effect of presence or absence of parents 
in the home; the effect of a large or small number of brothers 
and sisters, or none; the effect of married life on the scores of 
spouses; and the effect of children on the scores of their parents. 

THE INFLUENCE OF ORPHANAGE AND PART-ORPHANAGE 

There is a popular belief that the sons of widowed mothers 
tend to become feminized by the resulting relationship, and that 
daughters brought up chiefly by fathers tend to become mas- 
culinized. Table 52 presents the pertinent data on this point 
for our adult population of males and females. The mean 
scores of the various groups in the table are given separately for 
(a) the adult population of high-school education and (6) the 
total adult population. 

In both these populations, strange to say, the males brought 
up by widowed mothers average more masculine than those 
brought up by widowed fathers. The difference is approximately 
twice its standard error. Also, in both populations of males the 
wholly orphaned rate appreciably less masculine than the total 
population from which they are derived. Although the N'& 
involved in these comparisons are too small to make the differ- 

223 



224 SEX AND PERSONALITY 

TABLE 52. M-F SCORES or ORPHANS AND PART-ORPHANS 



Males, status 


Adults of high-school 
education 


Total adult group 


No. 
of 
cases 


Mean 
M-F 
score 


S.D.M 


No. 
of 
cases 


Mean 
M-F 
score 


S.D.M 


Care of mother chiefly or only 1 . . 
Care of father chiefly or only 1 . . . 
Part or whole orphan total .... 


25 
9 
43 
9 
344 


+63.3 
+46.1 
+52.4 
+28.3 
+45.4 


8.8 
14.9 
7.1 
11.5 
2.9 


85 
35 
149 
29 
1083 


+52.4 
+39.1 
+45.8 
+34.6 
+43.3 


4.82 
8 09 
3.97 
10.48 
1.57 


Whole orphan total 1 


Population from which derived . . 


Females, status 














Care of mother chiefly or only 1 . 
Care of father chiefly or only 1 . . . 
Part or whole orphan total 


72 
24 
113 
16 
820 


-74.8 
-75.3 
-77.1 
-93.2 
-84 7 


5.8 
7.4 
7.2 
11.2 
1.5 


187 
63 
303 
53 
1867 


-83.6 
-86 
-83.5 
-80.1 
-80 9 


3.28 
3 25 
2 44 
5.98 
1.00 


Whole orphan total 1 


Population from which derived. . 



1 "Only" if other parent died or was absent from before child's sixth birthday; "chiefly" 
if other parent died or was absent after child's sixth birthday and before twelfth. 
Without care of either parent from before sixth birthday. 

ences very reliable, our data may at least be interpreted in the 
negative sense as offering no support to the popular belief stated 
above. In the case of the females the differences, both in the 
high-school population and the total populations, are too small 
to be even suggestive of an influence. 

INFLUENCE OF SIBLINGS 

The possible influence of brothers and sisters on an individual's 
masculinity or femininity is a matter of widespread interest. 
One wishes to know whether masculinity or femininity as meas- 
ured by this test bears any relation to the number of siblings or 
to their sex. Does the only child tend to be more or less mascu- 
line than the average, and if such a difference obtains, does it 
hold equally for males and females? Does the lone boy with 
several sisters tend to become feminized, or the lone girl with 
several brothers masculinized? Table S3 for the males and 



RELATION OF M-F SCORE TO THE DOMESTIC MILIEU 225 



Table 54 for the females present mean M-F scores for groups 
having various types of sibling relationships. The data are for 
our adult population of high-school education. 

TABLE S3. INFLUENCE OF SIBLINGS ON THE M-F SCORE OF MEN OF HIGH-SCHOOL 

EDUCATION 



Sibling status 



No. of 
cases 



Mean 
M-F 
score 



S.D.jv 



Norm of adult population of high-school education with 
whom these groups, are compared 

a. Brothers, with or without sisters: 

Men who have brothers 

Men who have 1, 2, or 3 brothers 

Men who have 4 or more brothers 

b. Sisters, with or without brothers: 

Men who have sisters 

Men who have 1, 2, or 3 sisters 

Men who have 4 or more sisters 

c. Brothers in sibships without sisters: 

Men who have sisters, brothers 

Men who have sisters, 1 or more brothers 

Men who have sisters, 2 or more brothers 

Men who have sisters, 3 or more brothers 

Men who have sisters, 4 or more brothers 

d. Sisters in sibships without brothers: 

Men who have brothers, sisters 

Men who have brothers, 1 or more sisters 

Men who have brothers, 2 or more sisters 

Men who have brothers, 3 or more sisters 

Men who have brothers, 4 or more sisters 

e. Sibs (of either sex or both) : 

Men who have sibs 

Men who have 1 sib (either sex) 

Men who have 2 sibs (either sex or mixed) 

Men who have 2 sibs (mixed) 

Men who have 3 sibs (either sex or mixed) 

Men who have 3 sibs (mixed) 

Men who have 4 sibs or more (mixed) 

Men who have 6 sibs or more (either sex or mixed) 
Men who have 6 sibs or more (mixed) 



212 

50 

125 

35 

59 

130 

22 

19 
40 
21 
12 
6 

19 
31 
17 
10 

5 

19 

33 
36 
19 
27 
18 
85 
40 
39 



+50 

+54.9 
+55 6 
+ 18.5 

+52.9 
+51 6 
+23.2 

+42.1 
+57 
+61 
+38.8 
+33.8 

+42.1 
+62.7 
+56.4 
+50.5 
+42.5 

+42.1 
+61 4 
+73.8 
+68.4 
+42.4 
+42.7 
+41.1 
+22.0 
+21.8 



3.3 

6.1 
4.9 
8.9 

6.3 

4.7 

13.0 

4 8 

6.7 

12.7 

15 2 

12.8 

4.8 

7.7 

9.1 

12.6 

12.0 



13.3 

12.8 
6.1 
9.9 
9 5 



We may examine first section e of the above tables. Is there 
any relationship between M-F score and the size of the family 



226 



SEX AND PERSONALITY 



TABLE 54. INFLUENCE OF SIBLINGS ON THE M-F SCORE OF WOMEN OF 
HIGH-SCHOOL EDUCATION 



Sibling status 


No. 
of 
cases 


Mean 
M-F 
score 


S.D.M 


Norm of adult population of high-school education 
with whom these groups are compared 


820 


- 84.7 


1.5 


a. Brothers, with or without sisters: 
Women who have brothers 


121 


- 84.2 


3 7 


Women who have 1, 2, or 3 brothers 


332 


- 85.0 


2.4 


Women who have 4 or more brothers 


82 


- 84.9 


4 5 


b. Sisters, with or without brothers: 
Women who have sisters 


135 


- 81.1 


4 1 


Women who have 1, 2, or 3 sisters 


317 


- 85.9 


2.2 


Women who have 4 or more sisters 


83 


- 86.6 


5.2 


c. Brothers in sibships without sisters: 
Women who have sisters, brothers 


50 


- 82.7 


5.4 


Women who have sisters, 1 or more brothers . . 
Women who have sisters, 2 or more brothers . . 
Women who have sisters, 3 or more brothers . 
Women who have sisters, 4 or more brothers . 
d. Sisters in sibships without brothers: 
Women who have brothers, sisters . . . . 


85 
37 
19 
6 

50 


- 80.3 
- 89 5 
- 90.6 
-116.2 

- 82 7 


5.6 
8.8 
10.4 
17.4 

5 4 


Women who have brothers, 1 or more sisters . 
Women who have brothers, 2 or more sisters. . 
Women who have brothers, 3 or more sisters . . 
Women who have brothers, 4 or more sisters . . 
e. Sibs (of either sex or both) : 
Women who have sibs 


71 
39 
15 
9 

51 


- 85 6 
- 85 9 
-100.2 
- 96.2 

- 82 4 


4.8 
7.1 
10.0 
12.2 


Women who have 1 sib (either sex) 


79 


- 77 8 




Women who have 2 sibs (either sex or mixed) . . 
Women who have 2 sibs (mixed) 


83 
41 


- 81.8 
- 81 7 


6.4 


Women who have 3 sibs (either sex or mixed) . . . 
Women who have 3 sibs (mixed) 


77 
58 


- 85 6 
- 85 


5 1 


Women who have 4 sibs or more (mixed) 


229 


- 88 


2.7 


Women who have 6 sibs or more (either sex or 
mixed) . 


118 


- 87 


4.0 


Women who have 6 sibs or more (mixed) . . 


115 


- 86 9 


4 1 



from which an individual comes? The figures show a definite 
tendency for men from small sibships to score more masculine 
than the average, and for men of large sibships to score more 
feminine than the average. The correlation between an individ- 
ual's M-F score and the size of sibship is in fact .21 .04 for 



RELATION OF Jlf-F SCORE TO THE DOMESTIC MILIEU 227 

men and .35 .03 for women. One must be cautious, how- 
ever, in the interpretation of these figures. We have shown in 
an earlier chapter that M-F score correlated negatively with 
age, particularly with males, and the question arises whether our 
subjects from large sibships may not be older than those from 
small families. It turns out that the correlation between age 
and size of sibship is +.29 .04 for men and +.13 .06 for 
women. It seems, therefore, that age accounts for the negative 
correlation between M-F score and size of sibship in the case of 
the males, and to some extent in the case of the females. 

The relation of M-F score to number of brothers has been 
examined without reference to the presence or absence of sisters. 
In the case of men, those who have four or more brothers score 
more feminine than those who have three or fewer, a difference 
which may be due entirely to age. In the case of women the 
same trend is evident, but in reduced magnitude, and can 
again be attributed to the age factor. A similar comparison of 
the relation between M-F score and number of sisters, without 
regard to presence or absence of brothers, leads to the same 
general conclusion. The more feminine M-F score of men with 
four or more sisters, as compared with that of men with fewer 
sisters, probably represents an age effect. The corresponding 
data for women suggest that those with a large number of 
sisters tend to score just more feminine than the average for the 
entire population, and those with no sisters just more masculine. 
Again the influence is at least partly one of age. 

We may note next the influence of brothers in families in which 
there are no sisters. The figures for men are inconsistent. 
Those who have neither brothers nor sisters score just more 
feminine than the mean for adult men of high-school education. 
Those who have one or more brothers but no sisters score more 
masculine than the general mean. Neither difference is statis- 
tically significant. Where there are three brothers or more and 
no sister the relationship is reversed, perhaps because of the 
negative correlation between age and size of family. In the case 
of women a more regular trend is observed. Those with one 
brother and no sister score somewhat more masculine than the 



228 SEX AND PERSONALITY 

general population of which they are a part, and also more mascu- 
line than the similar subgroup having neither brothers nor 
sisters. From this point on there seems to be a definite tendency 
for each additional brother to have a feminizing influence. The 
differences all lack statistical significance if considered individ- 
ually, but their trend is constant. The safest conclusion seems 
to be the negative one that no evidence is found to support the 
popular view that girls who have many brothers but no sisters 
are masculinized, or that the reverse effect is very marked in the 
case of boys. 

Examination of the mixed sibships (Section e of Tables S3 and 
54) discloses no relationship of importance except the usual 
negative correlation of M-F score and size of family. This we 
have attributed to the effect of age. An interesting fact appears, 
however, when we compare for either sex the scores in Sections 
c and d of the tables. Here we find a tendency for men with 
siblings of one sex only to be more masculine if the siblings are 
sisters than if they are brothers. This suggests a contrast effect. 
When a man has both brothers and sisters the contrast effect is 
not evident. Women with siblings of one sex only tend to be 
more feminine if the siblings are sisters than if they are brothers. 
This is the opposite of the contrast effect observed for the men. 

RELATION OF THE M-F SCORE TO MARITAL STATUS 

Two interesting questions arise in this connection: (1) Is there 
marital selection on the basis of resemblance or difference in 
mental masculinity? (2) Does the intimate association of a man 
and a woman in married life exert a feminizing effect upon the 
husband and perhaps a masculinizing effect upon the wife, 
leading to a gradual rapprochement with respect to M-F score ? A 
group of 179 married couples was tested in an attempt to answer 
these questions. The husbands of the group included 83 in 
the teaching profession (all grades), 9 each in engineering and 
the ministry, 4 in medicine, 17 in business, 27 in clerical and 
mercantile occupations, and 21 miscellaneous. They represented 
all grades of education but half or more were college graduates. 



RELATION OF M-F SCORE TO THE DOMESTIC MILIEU 229 

The average age of the husbands was 34.2 years, that of the 
wives 33.3. The range of ages was from the early twenties to 
more than sixty. The average number of years married was 9.4. 

The mean for the husbands is +44.4, which is just below the 
mean of the male teaching groups in our general population. 
The mean for the wives is 76.7, or appreciably more feminine 
than the female teaching group in the general population. 
Husbands' scores correlated .25 .046 with number of years 
married, and wives' scores .19 .048 with number of years 
married. Taken at their face value these correlations would 
suggest a feminizing influence of marriage upon both husband 
and wife, but such an interpretation is unwarranted. The rela- 
tionship noted reflects chiefly the negative correlation between 
M-F score and age. When the difference between the score of 
husband and wife was correlated with length of marriage, no 
relationship was found (r = .026 .050). 

Correlation of M-F score with number of years married was 
also computed for a separate group of 161 married men and 
another of 402 married women, both groups of fairly homo- 
geneous education. For these men M-F scores gave a correlation 
of .28 .05 with number of years married; for the women 
.07 + .03. But the number of years married correlates with 
age .89 .01 for the men and .86 .01 for the women. With 
age constant the negative correlation between M-F score and 
length of married life disappears. There is accordingly no evi- 
dence that length of married life per se has any appreciable mas- 
culinizing or feminizing effect on either husband or wife. 

Turning to the question of marital selection, we find in the 
group of 179 couples a husband-wife correlation of .16 .05. 
It is .16 .07 for 89 couples married seven years or less and 
.20 .07 for 90 couples married eight years or more. In the 
separate exercises of the M-F test the husband-wife correlations 
are less than twice their probable error for all but two: Exercise 
3 (information) and Exercise 4 (emotional attitudes). For these 
the correlations were, respectively, .26 .05 and .31 .05. It 
is probable that the husband-wife correlations of .16 to .20 in 
total score reflect merely the marital selection that is known to 



230 SEX AND PERSONALITY 

occur with respect to age and amount of education and does not 
indicate homology based upon mental masculinity or femininity. 

EFFECT OF MARRIAGE ON M-F SCORES OF PROFESSIONAL AND 
OTHER WOMEN GAINFULLY EMPLOYED 

It was noted in an earlier chapter that women working outside 
the home tended to score more masculine than comparable 
housewives. In the general population group the mean for 
married women working was 73.8, for single women working 
74.8, for housewives with no other occupation 81.6, and for 
housewives who formerly worked 83.9. These and other data 
are brought together in Tables 55 and 56. 

The contrast in score in the tables is clearly not between 
married and unmarried primarily, but rather between working 
and not working. The interpretation might nevertheless be sug- 
gested by the figures that it is in general the more feminine 
women who marry (M-F score 80.7 versus 74.8 for the single 
women) and that among the married are a less feminine group 
(M-F score 73.8) who work after marriage. Analysis of the 
scores of the four categories decade by decade shows that this 
would be a mistaken interpretation based on figures weighted 
unevenly at different ages. Equal weighting for the decades 
changes the order of the categories (least to most feminine) to 
the following: married and working, housewives, housewives who 
formerly worked, single and working. Equal decade and cate- 
gory weighting gives an average M-F score of 87.4 (s.s. .05) 
for the single women and 83.9 (s.s.O) for the married 
women. In each decade the married women tend, of course, 
to be the older, hence the difference is not due to age. It seems 
from this comparison that mental masculinity is not a factor 
unfavorable to matrimony. However, comparison of the scores 
of married and single, with age and education made constant, 
has not been attempted. It seems at present from the figures 
available unlikely that such analysis would prove more illumi- 
nating than the material already available. A final comparison 
of the means for the married and single groups from age twenty to 
age sixty with equal decade and category weighting gives prac- 



RELATION OF M-F SCORE TO THE DOMESTIC MILIEU 231 



TABLE 55. M-F SCORES or ADULT WOMEN IN RELATION TO MARRIAGE AND 

OUTSIDE WORK 



Group categories 


Age 


No. of 
cases 


M-F 

score 


Single and working 


20-29 


239 


72 5 


Married and working 




92 


71 5 


Mamed were formerly working 




91 


84 2 


Married, have had no other occupation than that of housewife 


30-39 


74 
64 


-82.5 
80 1 






96 


82 2 


Married, were formerly working 




147 
84 


-86.5 
92 I 


Single and working 


40-49 


35 
46 


-87.8 
-84 7 


Married, were formerly working 
Mamed housewives . 




81 
77 


-79.1 
78 3 


Single and working 


50-59 


18 


91 7 


Married and working .... 
Mamed, were formerly working 
Married, housewives .... 
Single and working . . . 
Mamed and working . . 
Married, were formerly working 


60-69 


30 
60 
55 
11 
13 
33 
34 


-84.2 
-87.8 
-84.8 
-93.1 
-92.6 
-97.4 
-79 5 




70-79 


4 


-99 5 


Mamed and working 
Marned, were formerly working. 
Married, housewives 
Single and working 
Marned and working 


Total 


5 
12 
7 
371 
282 
453 


-65.5 
-91.2 
-86 6 
-74.8 
-73.8 
-83 9 


Mamed, have had no other occupation than that of housewife 
Equal decade weighting: 
Single and working 
Marned and working ... 
Marned, were formerly working 
Married, housewives 
Equal decade and category weighting: 
Single 




332 

371 
282 
453 
332 

371 


-81.6 

-87.4 
-80.1 
-87.7 
-83.9 

-87.4 






1067 


-83.9 


Equal decade and category weighting, ages 20 to 60. 




371 


-83.0 






1067 


83.1 


Grand total: 




371 


-74.8 


Married 




1067 


-80 7 











tically identical scores (M-F score 83). Summarizing all our 
results so far as they relate to this matter, it appears that with 
marriage (1) the better educated and widely experienced woman 
will show a slightly more feminine score as her interests become 



232 



SEX AND PERSONALITY 



I 

w 



o 

I 

i 



S 
o 

H ; 



|! 

5^ 

58 
Si 

i 
g 





s 







3 



1 



i 



i 



I-& 



I 



oo oo ON so 

l*- I s * OO OO 

I I I I 



c* ^ Q ^n 

\O si O VO 

1 1 1 1 



ON to ON *<f 

3 58 53 8 
I l I l 



\O ^ ^H 

ON VO i 



oo oo ON ON 
I I I I 



cj r*i ^p ro 

oo t> ON oo 

I I I I 



ON o OO 
r^ <v T^ 

O ON r>. 



co vO ON 

' v_x >: vix 

oo O O 




O 10 10 O 
o -^ 10 to 



10 O O 

9 o -^ -: 



* 1 OO ^H 



to 10 O >o 
O ^H 



ON to oo r- 

10 O ^o 

r + r 



CO -^ OO 

O o 




RELATION OF M-F SCORE TO THE DOMESTIC MILIEU 233 

more domestic; (2) this tendency will be more or less counter- 
acted if she continues her extradomestic activities; (3) the less 
well educated woman of narrow experience will become slightly 
more masculine in score, as marriage affords her wider experi- 
ential contacts. The influence of other factors such as education 
and intelligence are prepotent and are not overridden by the 
influence of marriage. 

The unmarried male population available for study is prac- 
tically limited to men under thirty-five years of age. Compari- 
son of the unmarried men (M-F score +60.1) with the married 
men (M-F score +56.2) from the group of high-school education 
and of the same age range shows the single men to be the more 
masculine, but the difference is less than half its standard error. 
The difference between the same population in its twenties and 
thirties is larger than this, and the difference between men who 
care much and men who care little for social life, irrespective of 
whether they are married, is almost twice as large. 

RELATION BETWEEN OFFSPRING AND M-F SCORES OF PARENTS 

Data on the relationship between number and sex of offspring 
and M-F scores of parents are presented in Tables 57 and 58. 

Married men and married women without children rate more 
masculine than the general mean. The unmarried of either sex 
are just more masculine than the mean of the same sex but less 
masculine than the married with no children. There is a slight 
negative correlation between number of children without regard 
to sex and M-F score for the men; the women show no correla- 
tion or a very slight one if one very masculine mother of four 
boys is omitted. The general trend is in line with the previously 
demonstrated correlation between age and score, for the number 
of children is of course correlated with age. The fact that the 
women show so slight a correlation, if any, between score and 
number of children seems to indicate that the children have a 
small but real influence which counterbalances the age effect. 

Comparison of the ranges and the trends indicated by the 
scores of the men who are fathers of girls only, or fathers of boys 
only, suggests that there may be a small influence of children 



234 



SEX AND PERSONALITY 



upon parents which is differential for sex of offspring. The 
fathers of girls only are increasingly feminine as the number of 
girls increases. Indeed the four fathers of three or more girls 
rate as feminine as sixty-year-old men. 

TABLE 57. RELATIONSHIP BETWEEN OFFSPRING AND M-F SCORE OF 

MOTHER 



Group categories 


No. of 
cases 


Mean 
M-F score 


S.D.M 


Norm for women of high-school education (Series 7) 
Unmarried women 


536 
124 


- 84.1 

- 83.5 


2.3 
4.0 


Married women, children 


109 


- 79.4 


4.4 


1- Number of children and M-F score of mother: 
Mothers of 1 child (male) 


51 


- 84.8 


5.7 


Mothers of 1 child (female) 


48 


- 90.8 


5.3 


Mothers of 1 child (either male or female) 


99 


- 87 7 


6 3 


Mothers of 2 children (mixed) 


41 


- 91.0 


4 9 


Mothers of 3 children (mixed) 


32 


- 86.4 


7.1 


Mothers of 4 children (mixed) 


30 


- 94 8 


8.0 


Mothers of 5 children (mixed) 
Mothers of 6 or more children (mixed) 


18 
22 


- 76.2 
- 85.0 


11.1 
7.0 


2. Number of boys in families with boys only 
and M-F score of mother: 
Mothers of girls, 1 boy 


51 


- 84 8 


5.7 


Mothers of girls, 2 boys 


18 


- 87.3 


10 2 


Mothers of girls, 3 boys ... 


9 


- 93 9 




Mothers of girls, 5 boys . . 


1 


- 30.0 




3 . Number of girls in families with girls only 
and M-F score of mother: 
Mothers of boys, 1 girl 


48 


- 90.8 


5.3 


Mothers of boys, 2 girls 


26 


- 88.0 


8 7 


Mothers of boys, 3 or 4 girls 


7 


-100 9 


16 6 



A father who has one boy is more feminine than the unmarried, 
the married with no children, or the general average. But he is 
more masculine than the father of one girl, and the addition to 
his family of successive boys and no girls seems to make him 
increasingly more masculine. The five fathers of five or more 
boys and no girls are more masculine than the unmarried and 
almost reach the high point of the married with no children. 
Although undoubtedly much older than the average of either 
of the other two groups mentioned, they exceed even the twenty- 



RELATION OF M-F SCORE TO THE DOMESTIC MILIEU 235 



and thirty-year age norms for their own or any other group. It 
seems strongly suggested therefore that the influence of boys 
on the father is definitely positive and would be found quite 
marked were the contraactive factor of age eliminated. 

TABLE 58. RELATIONSHIP BETWEEN OFFSPRING AND M-F SCORES OF 

FATHERS 



Group categories 


No. of 
cases 


Mean 
M-F score 


S.D.M 


Norm for men of high-school education (Series 7) ... 
Unmarried men 


212 

48 


4- 50.0 
4- 60.1 


3.3 
7.6 


Married men, children 


39 


4- 67.4 


9.1 


1. Number of children and M-F score of father: 
Fathers of 1 child (male) 


24 


-1- 41 3 


11 8 


Fathers of 1 child (female) 


16 


4- 39 2 


11 9 


Fathers of 1 child (either male or female) 
Fathers of 2 children (mixed) 


40 
20 


4- 40.5 

4- 58 5 


11.9 
11 


Fathers of 3 children (mixed) 


14 


4- 46 2 


9.7 


Fathers of 4 children (mixed) 


8 


4- 23.0 


13.2 


Fathers of 5 children (mixed) . . . 


7 


4- 56.2 


11.2 


Fathers of 6 or more children (mixed) 


3 


4- 33.7 


5.4 


2. Number of boys in families with boys only and 
M-F score of father: 
Fathers of girls, 1 boy 


24 


4- 41.3 


11.8 


Fathers of girls, 2 boys 


7 


4- 53.4 


11.7 


Fathers of girls, 3 or 4 boys . ... 


5 


4- 66.5 


6.7 


Fathers of girls, 5 boys 


1 


4-130.0 




3. Number of girls in families with girls only 
and M-F score of father: 
Fathers of boys, 1 girl 


16 


4- 39.2 


11.9 


Fathers of boys, 2 girls 


14 


4- 26.2 


14.0 


Fathers of boys, 3 or 4 girls 


4 


4- 20.5 


20.6 











A difference in the influence exercised by male children as 
compared to that of female children is also to some extent 
apparent among the women. The mothers of boys only are 
more feminine than the general mean, than the unmarried, and 
than the married with no children. But the differences are very 
slight and the range from the 51 mothers who have each one boy 
only to the 27 who have two or more is only .10 s.s. This is 
certainly not more and may be less than the expected correlation 
with age. The mothers with girls only are more feminine 



236 SEX AND PERSONALITY 

than the mothers with boys only. The range from the 81 
mothers who have one daughter each and no sons to the 7 mothers 
who have three or more daughters each and no sons is .25 s.s. 
The 7 mothers with three daughters each and no sons score 
considerably below any of the decade norms. There is at least 
a suggestion that the sex of the child may have a real influence 
upon the M-F score of the mother. 

Comparing the two parents, similarity and also difference 
in the effect of children upon their scores appear. Both fathers 
and mothers are feminized, though very slightly, by the presence 
of a single child without regard to its sex. Both are feminized 
by the presence of daughters in proportion to the number of the 
daughters. The trend is probably in addition to the f eminization 
expected with age. Both fathers and mothers are masculinized 
by the presence of sons. The fathers show actually more 
masculine scores with an increasing number of sons, an effect 
the more significant since it is the opposite of that expected with 
increasing age. Mothers show almost complete counteracting 
of the feminization of age as the number of sons increases. 
Perhaps we may say that fathers are more diversely affected by 
sons and daughters, mothers are more similarly affected. 

A word may be added regarding the M-F scores of the unmar- 
ried and the married with no children. The unmarried of both 
sexes are slightly more masculine than the general means. This 
should doubtless be attributed chiefly to the fact of their rela- 
tively greater youth. In each sex the married without children 
are, however, and strangely, more masculine than the unmarried 
although they are of course older. The difference between 
the mean for the married men with no children and the 
general mean is more than three times its standard error. The 
difference between the mean for the married men with no children 
and those with one or more daughters is also more than three 
times its standard error. 

Most of the differences discussed in this section are in large 
part associated with age differences. In so far as age is not the 
conditioning factor they are to be regarded as suggestive only, 
since for these small groups the significance of the differences 



RELATION OF M-F SCORE TO THE DOMESTIC MILIEU 237 



is, except in the few instances mentioned, not established. 
Where regular trends have appeared and reappeared consistently 
the possibility of their genuineness should, however, not be 
denied without the evidence of further data. 

FAMILY RESEMBLANCES IN M-F SCORES 

Family resemblances in mental masculinity and femininity 
have not been extensively investigated. In Chapter VI data 
were presented showing that on a modified M-F test (based upon 
Strong's vocational-interest test) the correlation between 
husband and wife is practically zero. In Chapter X the corre- 
lation for M-F scores of 179 husbands and wives was found 
to be slight but perhaps significant. We have found consider- 
able evidence in case history data that the girl who tests extremely 
masculine is likely to have a father with strong mechanical 
interests, and that the sons of very religious mothers are likely 
to test rather feminine, but it remains to be seen whether these 
indications will be confirmed by statistical evidence. 

The only parent-offspring correlations we have are for small 
populations of gifted children (IQ 140 or higher). These are as 
follows: 



Groups compared 


N 


r 


Sons with mothers . 


46 


138 


Sons with fathers. . . . 


31 


168 


Daughters with mothers 


39 


.009 


Daughters with fathers . 


28 


428 


Boys with parental mean 


26 


413 


Girls with parental mean 


24 


179 



As the probable errors range from .10 to .12, only two of the above 
correlations are statistically significant, and there is a small 
chance that these also are caused by sampling errors. The 
problem would repay further investigation. It would be 
especially desirable to establish parent-child resemblance in 
M-F score in comparison with the resemblance between foster 
parents and their foster children adopted in infancy. 



238 SEX AND PERSONALITY 

SUMMARY 

1. Part orphans tend to be contrastingly masculine or femi- 
nine in mental traits according as they are brought up by the 
mother or the father. Whole orphans are if anything more 
feminine than the average. 

2. The brother of one sister tends to be more masculine than 
the average, and the sister of one brother tends also to be more 
masculine than other women. Aside from these trends the 
influence of siblings is not to be distinguished from the influence 
of age, for the size of sibship is correlated with the age of the 
individual and the correlation of the latter with the M-F score 
probably explains the similar correlation of the former. 

3. A negative correlation demonstrated between length of 
married life and M-F score is found to persist only when a wide 
age range is considered. When the age factor is held constant 
the correlation disappears. Single and married women rate 
at the mean when other factors are weighted as nearly equally as 
possible. 

4. Children appear to influence the M-F scores of parents. 
The fathers of one child are more feminine than the mean and 
than married men with no children. The latter are more 
masculine than the unmarried. With increasing size of family 
both father and mother tend to score more feminine. A single 
exception is found in the case of fathers of boys only. These 
fathers are increasingly masculine as the number of boys increases. 
Mothers show a similar trend in the retardation of feminine 
emphasis which occurs when the children are all boys. 



CHAPTER XI 
A STUDY OF MALE HOMOSEXUALS 1 

A study of male sexual inverts was undertaken in the hope that 
it would contribute information of value in the interpretation of 
M-F scores and possibly throw some light on the genesis of 
homosexuality in males. It was entered upon with little first- 
hand knowledge about the personality traits characteristic of 
such a group, and with no preconceptions in regard to the causes 
of sexual inversion. It is believed that the results of the study 
throw a certain amount of light on both of these questions. 

SUBJECTS TESTED 

Because homosexual persons in the general population live in 
constant fear of public exposure and consequent social ostracism, 
they are not easily accessible for study. Jails, prisons, and 
reform schools offer the best opportunity for obtaining subjects. 
A number of such institutions were searched for possible cases 
with the results shown below. One of the subjects found in the 
San Francisco County Jail agreed to cooperate in making con- 
tacts with additional cases among his associates. By his help 
and that of others secured through him a fairly large number of 
noninstitutional subjects have been found. The facts con- 
cerning the number of subjects, where found, and the type of 
homosexuality represented by each group are shown in Table 59. 

A few definitions of the terms used in this report should 
perhaps be given. A homosexual is a person who out of prefer- 
ence has sexual relationships with persons of the same rather than 
of the opposite sex. Since this study deals with male subjects, 
the term male homosexual refers to a man who has sexual 

1 The data of this chapter were collected and treated entirely by Dr. E. Lowell 
Kelly under the general direction of Terman and Miles. This summary was 
prepared by Dr. Kelly and Dr. Terman. 

239 



240 



SEX AND PERSONALITY 



TABLE 59. DATA CONCERNING THE NUMBER or SUBJECTS, WHERE FOUND, AND 

TYPE OF HOMOSEXUALITY REPRESENTED BY THE GROUPS 

STUDIED 



Location of Subjects 


No. 


Classification 


State Prison at San Quentin 


5 


Passive 


State Prison at San Quentin 


6 


Doubtful 


Berkeley Police Department . 


2 


Passive 


San Francisco County Jail, No. 2. . 


11 


Passive 


San Francisco County Jail, No. 2. . . 


s 


Doubtful 


General population . . 


S9 


Passive 


U.S. Disciplinary Barracks at Alcatraz, California. . . . 


46 


Active 


Total in each classification . . . 


46 
77 
11 


Active 
Passive 
Doubtful 


Grand total 


134 


All types 









relationships with men rather than with women. He is said to 
be an active homosexual if he plays the male role in the copulatory 
act, and a passive homosexual if he plays the role of female. 
These relationships may be further classified according to the 
particular type of sexual connection practiced. Rosanoff 1 lists 
six general types: 

1. Mutual masturbation 

2. Interfemoral coitus 

3. Irmmation 

4. Mutual irrumation 

5. Sodomy 

6. Combinations of these 

Irrumation, or fellatio, refers to copulation of penis with the 
mouth. Sodomy, used synonymously with pederasty, refers to 
copulation involving the penis and anus. These terms are 
rarely if ever used by the homosexuals as they have an extensive 
slang vocabulary of their own. 

In this study a subject is called passive if he currently plays the 
female role in either fellatio or sodomy. A total of 77 such cases 



1 ROSANOFF, AARON J., Manual of psychiatry, p. 202, Wiley, 1920. 



A STUDY OF MALE HOMOSEXUALS 



241 



have been given the M-F test. The 46 army prisoners were all 
serving sentences for sodomy. Captain G. E. Hesner, staff 
psychiatrist at the time, reported that in practically every case 
the prisoners in question were supposed to have played the active 
part, and the group has therefore been classified accordingly. 
Those classified as doubtful were persons suspected of homosexual 
acts, but for whom there was no information as to the role played. 

ANALYSIS OF M-F SCORES 

The M-F test was given to each of the 134 cases. Form A was 
used for all except the Alcatraz group, who were given Form B. 



i i t i i i 

LEGEND 

""77 Passive homosexuals 
98 High school boys 




-160 -140 -120-100 -80 -60 -40 -20 +20 *40 +60 +80 +100 +120 +140* 160 +160 +200 
to to to to to to to to to to to to to to to to to to to 
179-159 -139-119 '99 "79 '59 -39 -19 +1 *21 +41 +51 +81 +101*121*141+161+181 

M-F Scale 

FIG. 14. Distributions of M-F scores of 77 passive male homosexuals, 98 high- 
school boys, and 90 high-school girls. 

Since the two forms of the test are fairly comparable, the raw 
scores have been used in all computations. 

The distribution of the total scores made by the 77 passive 
homosexuals, comprising the largest group, hereafter referred 
to as the P.M.H. group, is shown in Fig. 14, together with the 
distributions of scores for unselected groups of third-year high- 
school boys and girls. 1 It may be seen that the curve for the 
P.M.H. group approximates more closely that for the girls than 
that for the boys. However, it is distinctly different from that of 

1 The distribution of numerical scores of the P.M.H. group (including three 
additional cases) is given in Table 10. 



242 



SEX AND PERSONALITY 



either, the difference between the means in each case being 
statistically significant beyond question. 

No other male group thus far studied has tested anything like 
as feminine as these 77 passive male homosexuals. The male 
groups nearest to them in mean M-F score are the artists, 
ministers, Protestant theological students, priests in training, and 
unselected men in the seventies and eighties. In fact, the mean 
score of the P.M.H. group (19.7) is more feminine than that of 




FIG. 



-160 -WO -120 -KW -80 -60 -40 -20 +20 +40 +60 +80 +100 120*140 160180*200 
to to to to to to to to to to to to to to to to to to to 
-179 -159 -139-110 -99 "79 -59 "39 -19 + 1 +21 +41 +61 +81 *101 +121 +141 +161 +181 

M-F Scale 

IS. Distributions of M-F scores of 77 passive male homosexuals, 46 active 
male homosexuals, and 42 regular army men. 



38 college women athletes (13.7) and much more feminine 
than that of a small subgroup of these athletes who rated high 
in intelligence and low in scholarship (see pp. US/, and Fig. 6). 
Figure IS gives the score distribution of the 77 passive homo- 
sexuals, together with the distribution for the 46 Alcatraz 
prisoners classified as active, and the distribution for 42 unselected 
soldiers in the regular army. Although the classification of the 
entire Alcatraz group as actives involves a certain amount of 
error, it will be seen that this group not only tests much more 
masculine than the passive homosexuals, but also significantly 
more masculine than the army group of roughly comparable 
ages. If these two groups of homosexuals are fair samples of the 



A STUDY OF MALE HOMOSEXUALS 



243 



active and passive types, it would appear that they in fact 
present a true dichotomy of personality. It is probable that the 
small amount of overlap of the two distributions is caused by the 
errors of classification which must have occurred, particularly 
in the group classified as active. The distribution of scores for 
this group is distinctly bimodal, indicating that the group 
contains two essentially contrasting populations. 1 

Do passive male homosexuals give a preponderance of feminine 
responses in all the exercises of the M-F test, or only on a part of 
them? Table 60 gives the mean on each of the seven exercises 
for the P.M.H. group and for 98 third-year high-school boys, the 
differences between the means of the two groups, and the ratios 
of the differences to their standard errors. 

TABLE 60. COMPARISON OF MEAN SCORES OF 77 PASSIVE MALE HOMOSEXUALS 

AND 98 HiGH-SCHOOL-JUNIOR BOYS ON THE INDIVIDUAL EXERCISES OF 

THE M-F TEST 



Exer- 


Homosexuals 


H.S. Boys 


Diff. 




Diff. 


cise 


Mean 


S.D.a,, 


Mean 


S.D* 


Means 




S.D.diff. 


1 


- 5.72 


8.26 


- 93 


8 02 


4 79 


1.26 


3.80 


2 


- 0.86 


1 29 


- 0.27 


0.56 


0.59 


.16 


3.69 


3 


- 0.08 


8.89 


+ 6 68 


8 06 


6.76 


1.30 


5.20 


4 


+21.26 


27.26 


+ 15 19 


24.16 


6.07 


3.94 


1.54 


5 


-36.16 


30.92 


+49.26 


31.32 


85.42 


4.73 


18.10 


6 


- 2.84 


5.63 


+ 1.72 


8.36 


4.56 


1.06 


4.30 


7 


- 2.70 


2.51 


- 0.09 


1 75 


2.61 


.34 


7.68 


Total 


-27.97 


42.25 


+71.52 


51.96 


99.49 


7.08 


14.05 



Table 60 shows that the P.M.H. group tested feminine on all 
the exercises of the test except 4 (emotional response), in which 
it was slightly more masculine than the high-school boys. This 
reversal of the general trend is probably to be interpreted as a 
lessening of emotional and moral sensitivity due in part to the 

1 A group of 9 homosexual boys was tested at the Preston School of Industry 
(California). These included one classified by the school officers as active who 
tested at +56, and eight classified as passive whose scores were: 104, 66, 
-49, -43, -38, -2, +59, +63. The last two were in all probability active 
but wrongly classified. 



244 SEX AND PERSONALITY 

greater age of the P.M.H. subjects, but perhaps more especially 
to their associations and the type of life they lead. The greatest 
difference between the two groups is on Exercise 5, dealing with 
interests, likes, and dislikes. Here the P.M.H. mean (36.16) 
is practically identical with that of third-year high-school girls 
(-36.40). 

Can we accept the feminine scores of the P.M.H. group as 
reflecting the true personalities of the subjects their actual 
interests and attitudes or do they reflect primarily a conscious 
effort of the subjects to ape the feminine pattern just as they are 
known to do in physical appearance, carriage, and speech? It 
has been shown (Chapter IV) that the average man, knowing 
the intent of the test, can by taking thought score even more 
feminine than the average woman. These subjects did not 
know the exact purpose of the test, but many of them knew and 
others may have guessed that they were being studied because 
of their homosexuality. Perhaps this knowledge was stimulus 
enough to make them consciously try to respond as femininely 
as possible; or if not consciously, perhaps unconsciously, with a 
result no less misleading than if the score had been purposely 
faked. 

There are several reasons why we do not believe that this 
theory can account for our findings. When reform school 
youths or prison groups are tested en masse without any of them 
knowing or even suspecting the purpose of the test, we still find 
that those known by the officials to be passive homosexuals 
nearly all score well down in the feminine range, while those 
known to be active homosexuals usually score well up in the 
masculine range. We tested a number of individual homosexual 
subjects who had been completely misled as to the purpose of the 
test and had no suspicion that they were known by the examiner 
to be sex inverts, and in these cases too the characteristic type 
of score is almost invariably found: a feminine score for the 
passive male homosexual and a masculine score for the active 
male homosexual. 

Even more convincing evidence that the scores of the P.M.H. 
subjects are not faked, either consciously or unconsciously, is 



A STUDY OF MALE HOMOSEXUALS 245 

found by comparing their scores in the separate exercises of the 
test with those of young men who have been instructed to respond 
to all the items as a very feminine woman would respond. Under 
these conditions the score becomes more feminine in all the seven 
exercises, but to an unequal degree. As we go from exercise to 
exercise, the mean change it is possible for a group to make 
describes an irregular profile. On Exercises 2, 3, and 6 (ink 
blots, information, and historical characters) the faked scores 
are only a little more feminine than the naive scores; on Exer- 
cises 1 and 7 the change is greater but still moderate; and Exer- 
cises 4 and 5 (emotional response and interests) are alike in 
showing changes of far greater magnitude. In other words, 
males are able to guess better in some fields than others the way 
in which responses of typical women would differ from their 
own. They guess best with respect to interests and preferences, 
and with respect to emotional responses of anger, fear, pity, 
disgust, and ethical appraisal. 

When we examine the divergences of our P.M.H. subjects from 
normals on the seven exercises we find that they make a profile 
pattern decidedly unlike that of the intentionally faked scores. 
For example, Exercise 4, in which it is easy to fake the score of the 
opposite sex, rates the P.M.H. group even less feminine than 
the normal group. One may conclude, therefore, that the highly 
feminine scores of the P.M.H. group on Exercise 5 are genuine. 
There is no reason why subjects anxiously desiring to appear 
feminine should respond very femininely in Exercise 5 and 
more than normally masculine in Exercise 4. 

ANALYSIS or THE P.M.H. GROUP 

The average age of the 77 subjects constituting the P.M.H. 
group is 26.0 years. The distribution of ages has a range from 
17 to 44 years and a standard deviation of 6.12 years. It is 
skewed toward the younger ages, possibly due in part to the 
selection involved in obtaining the subjects. Since the subject 
who helped secure cases was only 22 years old, he probably 
tended to bring acquaintances who were somewhere near his 
own age. Another possible cause for the skewness may be 



246 SEX AND PERSONALITY 

found in the report that many of the older individuals of this 
type do not enter into the social life of the group and are therefore 
not available for study. They have generally better economic 
status and have more to lose through public exposure. 

Nine out of 70, or 12.9 per cent of the group for whom family 
data are available, were "only " children while 21, or 30 per cent, 
were firstborn. The average age of the fathers of the group 
at the time of the subjects' birth was 31.7, of the mothers 26.8, as 
compared with 33.6 and 29.0 respectively for the parents of 
643 gifted children. 1 The average number of brothers is 1.5 and 
of sisters also 1.5. With respect to the data just given, the 
P.M.H. group probably differs little from the generality of men 
of their age and social origin. 

Only 4, or 5.7 per cent, of this group were married, as compared 
with 54.2 per cent of the twenty-six-year-old males in the general 
population. 2 Six others, or 8.5 per cent, have been married 
but at the time of this study were divorced or separated from 
their wives. The entire group of 77 men records a total of only 
5 offspring. These figures may be influenced by various selective 
factors, but they are important in showing the tendency of the 
members of this particular group in regard to participation in 
family life. 

At the time they were tested, 6 of the 77 were in a state prison 
and 10 were in jail. Of these 16 institutional cases, 13 regularly 
practiced prostitution as a means of livelihood previous to 
incarceration. Of the 54 noninstitutional cases for whom data 
are available, only 10 were known prostitutes, but it is believed 
that the actual number engaged in this occupation was at least 
two or three times as great as the reported number. The current 
occupations reported and the number engaged in each for the 
group of 54 noninstitutional cases were as follows: clerks, 7; 
show workers, 5 ; waiters, 5 ; cooks, 5 ; bookkeepers, 2 ; hospital 
orderlies, 2; house-to-house salesmen, 2; nurses, 2; ranch hands, 
2; seamen, 2; high-school students, 2; and 1 each for bus boy, 

1 TERM AN, L. M., et al., Genetic studies of genius, vol. I, Mental and physical 
traits of a thousand gifted children, Stanford Univ. Press, 1925, Chapter VI. 
8 Census of the United States, 1920, vol. II, p. 391. 



A STUDY OF MALE HOMOSEXUALS 247 

artist, common laborer, cutter and pattern maker for ladies' 
dresses, porter, truck driver, bellhop, costume designer, house 
painter, bank cashier, fireman, janitor, ironworker, and 
schoolteacher. 

In order to ascertain whether or not the M-F scores of these 
77 subjects bore any relationship to the occupations followed 
by them, the following correlational study was carried out. Six 
judges 1 were asked to rate each of the above-mentioned 26 
occupations on a scale of 1 to 10 with respect to the masculin- 
ity-femininity of the work it involves. Although the mas- 
culinity-femininity of an occupation is a rather intangible 
characteristic, the fact that the intercorrelation of the ratings by 
the six judges were all between .75 and .92, with a mean of .84, 
indicates that the judges were fairly well agreed in -regard to the 
thing being rated. Their composite rating may therefore be 
regarded as a fairly reliable criterion of the masculinity-femininity 
of the occupations considered. 

These composite ratings were then correlated with the M-F 
scores of the 77 subjects. The correlation between them and the 
total M-F scores was found to be .24 + .07, indicating a very 
small but probably significant relationship. The correlation 
would certainly have been higher but for the fact that practically 
all of the occupations listed are those commonly associated with 
women and so have a restricted range. This fact in itself is 
significant and agrees with the data presented in Chapter VIII 
indicating that many occupations exert a selective influence 
which is reflected in the mean M-F scores of occupational groups. 

Approximately nine-tenths of the members of the homosexual 
group have had at least a grammar-school education, 11 are 
high-school graduates, and 6 have had one or more years in a 
college or university. On the whole, the group has a taste 
for good literature, music, and entertainment, if conver- 
sation can be taken as a reliable indication of actual interests 
and tastes. 



^he ratings were made by K. M. Cowdery, P. R. Farnsworth, Quinn 
McNemar, C. C. Miles, E. K. Strong, Jr., and L. M. Terman. 



248 SEX AND PERSONALITY 

At the time the study was made every subject in the group was 
accustomed to play the passive role in either fellatio or pederasty, 
and many engaged in both practices. The majority of them 
played only the passive role, although a few took the active 
part if occasion demanded. Practically all of them have lived 
with men in the relationship of man and wife. These unions are 
usually brief, although in a few cases they have lasted for as 
long as two or three years. In such cases, the active member 
of the pair frequently provided economic support for both. In 
a few instances the passive masqueraded as a woman and the 
two were legally married. 

FEMININE CHARACTERISTICS OF PASSIVE MALE HOMOSEXUALS 

The behavior of these male homosexuals, both as reported by 
them and as observed during the investigation, is as different 
from that of a group of average men as one could possibly 
imagine. It is well known that the average boy or young man 
makes every effort to keep from appearing effeminate. The 
passive male homosexual, on the contrary, takes advantage of 
every opportunity to make his behavior as much as possible like 
that of women. He not only accentuates any feminine qualities 
which he may already possess, such as a high-pitched voice, but 
also attempts to imitate women in his speech, walk, and manner- 
isms. Practically every subject has adopted a " queen " name 
by which he is known among his associates. They constantly 
refer to themselves as "the girls." Their behavior often seems 
exaggerated and ridiculous, although in some cases the inversion 
of behavior is remarkably complete. 

As was pointed out before, the group tends to follow no occcu- 
pation other than prostitution except occupations in which 
women or very effeminate men commonly find employment. On 
the whole they are far lazier than any persons of similar age with 
whom we have come in contact, being inclined to shun any 
occupation which promises even a small amount of hard work. 

Each subject in the P.M.H. group was asked to name his 
favorite study when in school, his favorite childhood play or 
game, and his favorite childhood toy or plaything. Replies were 



A STUDY OF MALE HOMOSEXUALS 249 

received by about 90 per cent of the group, many of the subjects 
naming more than one preference in one or more of the categories. 

The preferences for school studies were as follows, from most to 
least popular: history, 18; reading, 8; geography, 8; English, 5; 
spelling, 5; literature, 5; grammar, 5; mathematics, 5; art, 4; 
arithmetic, 4; foreign language, 4; drawing, 3; music, 3; and 1 
each for chemistry, philosophy, typing, and writing. Grouping 
into larger categories the subjects which are most alike we have: 
language studies, 33 ; history, 18; art (or drawing) and music, 10; 
arithmetic and mathematics, 9; geography, 4; miscellaneous, 4. 
On the whole the preferences are strikingly feminine; language 
studies, history, art, and music account for 61 of the 78 named. 

The plays and games preferred were as follows: playing house, 
17; games played with a ball, 12; marbles, 4; swimming, 3; tag, 
3; girls' games (not specified), 3; skating, 2; jumping, 2; and 
1 each for blocks, blind man's buff, checkers, chess, cro- 
quet, drawing, dressing up, dressing dolls, playing farm, fishing, 
hunting, hide-and-seek, hopscotch, hobbyhorse, jumping rope, 
music, ping-pong, post office, puzzles, reading, racing, playing 
store, theatricals, and tiddlywinks. The list of preferred play 
activities is even more feminine than the school-subject prefer- 
ences. Probably no other male group could be found which 
would rate "house" as its most preferred play. Except for 
"ball," nearly all of the activities named are more feminine than 
masculine, and there is such a variety of ball plays that we 
cannot know how many of the 12 naming this activity had in 
mind the more masculine types of ball play. Playing house is 
the third most feminine of 90 kinds of plays, games, and activities 
according to data collected by the senior author for a large 
group of unselected schoolchildren in grades 3 to 8. 

Applying the masculinity-femininity ratings of these 90 
activities 1 to the expressed preferences of the P.M.H. group 
(omitting the indefinite preference recorded as "ball"), the mean 
rating of the preferred activities is 10, on a scale in which 13 to 
20 is masculine and 12 to 1 is feminine. The mean would have 

1 See Table 1, p. 12. 



250 SEX AND PERSONALITY 

been still lower but for the fact that the subjects were asked to 
name their favorite "game." The connotation of the word game 
is such that it brought no mention of doll play and certain other 
kinds of feminine activities very popular in this group. 

Favorite playthings and toys are in the main such as would be 
expected from the activity preferences. The order is: dolls, 20; 
trains, 10; blocks, 5; bicycles, 4; books, 4; balls, 3; wagons, 3; 
teddy bears, 2; dishes, 2; marbles, 2; skates, 2; and 1 each for 
cats, dominoes, dresses, drums, engines, girls' toys (unspecified), 
hobbyhorse, mechanical toys, paints, piano, sailboat, tricycle, and 
top. It is profoundly significant that the plaything which has 
a two-to-one lead over its nearest competitor is perhaps the most 
characteristically feminine of all playthings known to Occidental 
culture. "Ball" receives only three mentions, which bears 
out the suspicion that some of those who named ball as a favorite 
game were not referring to the more masculine types of this 
activity. 

In weighing the significance of the preference reports one must 
of course bear in mind that they relate to experiences which are 
for the most part from ten to thirty years in the past (about 
twenty, on the average), and that the choices given are probably 
influenced by whatever femininity has been acquired since 
childhood. That the picture is probably in accord with the facts 
is, however, strongly suggested by the more detailed information 
secured in connection with the case-history studies summarized 
in Chapter XIII. It appears that the large majority of passive 
male homosexuals are characterized from fairly early childhood 
by an inversion of interests, attitudes, and activities. The 
"pansy" type of behavior of adult inverts is not primarily an 
affectation or the result of "abnormal" or "perverted" sexual 
practices. The explanation is to be sought either in terms of 
psychological conditioning or in terms of biochemical factors 
present from early childhood. If the causes were biochemical 
it would be reasonable to expect that they would affect physical 
development, especially the development of secondary sexual 
characteristics. We turn next to the evidence available for our 
P.M.H. group on this point. 



A STUDY OF MALE HOMOSEXUALS 251 

PHYSICAL TRAITS OF THE P.M.H. GROUP 

In order to determine the relation of physique to M-F scores 
and types of behavior, physical measurements were made on all 
members of the P.M.H. group whose cooperation could be 
secured. The data include measures (in centimeters) of total 
height, sitting height, suprasternal height, pubic crest height, 
shoulder diameter, hip diameter at the iliac crest and at the 
greater trochanter, knee width, hip circumference, and head 
circumference. They include further the approximate net 
weight in pounds, shoulder slope in degrees, and masculinity- 
femininity ratings, on a five-point scale, of the hair on the face, 
chest, legs, and pubic regions. Hair color, skin texture, and 
condition of genitals were noted. The measurements of height, 
sitting height, hip diameter, and shoulder diameter were made 
in the manner described by Dr. W. R. Miles 1 and with the aid 
of instruments supplied by him. Dr. Miles also gave his personal 
assistance in planning the techniques for the other measures 
listed. 2 

The measurements listed were secured for 33 cases, of whom 
24 were members of the P.M.H. group. The other 9 cases were 
obtained at the Preston (California) School of Industry, an 
institution for older juvenile delinquents. As these were con- 
siderably younger than the P.M.H. group their measurements 
were treated separately. The means and standard deviations 
of the measures for the group of 24 are given in Table 61 with 
comparative data when available for 100,000 white army troops, 
552 Stanford freshmen, and 100 Smith College women. Table 62 
gives the means and standard deviations for the small Preston 
group. 

Table 61 shows that in general the physical measurements of 
these male homosexuals do not differ markedly from those of 
army and college men, although there is a slight tendency for 

1 MILES, W. R., Human body weight: I, Correlations between body widths 
and other physical measurements. Science, 1928, 68, 382-386. 

1 Dr. Alvin J. Cox, Jr., Medical Assistant at St. Luke's Hospital, San Francisco, 
assisted Dr. Kelly in making the measurements. 



252 



SEX AND PERSONALITY 



TABLE 61. PHYSICAL MEASUREMENTS OF 24 HOMOSEXUAL MEN, 100,000 WHITE 
ARMY TROOPS, 552 STANFORD FRESHMEN, AND 100 SMITH COLLEGE 

WOMEN 



Item 


Homosexual men 


Army troops 1 


Stanford men* 


Smith women* 


Mean 


S.D.di.t. 


Mean 


S.D.dfat. 


Mean 


S.D. d kt 


Mean 


S.D. di .t. 


Age 


26.21 
170.75 
141.45 
88 9 
141.75 
86 5 
37 58 

28 44 

32 29 
111.3 
90 75 
56.83 
20 89 
1 33 


5.87 
7.67 
19 52 
2.96 
3.41 
5 87 
2 18 

1 50 

1 71 
4 92 
6.22 
1.89 
1.65 
.08 


171.99 
144.67 
90.39 
141.18 
86 82 

29 43 


6 66 
16 92 
3.51 
5.91 
5.05 

2 85 


19.65 
176.0 
141 3 
88 8 

37 4 
27 9 

1 34 


6*1 
16.2 
3 15 

2 1 
1 5 


20 38 
163 5 

86.29 

83 8 
35 1 

27 1 
1 30 


1 61 
5 21 

3.04 

4 08 
3 64 

1 93 


Height 


Weight 


Sitting height. . 
Sternal notch height 
Pubic height 
Shoulder diameter . 
Hip diameter, iliac 


Hip diameter, tro- 
chanter 


Shoulder slope 
Hip circumference 
Head circumference 
Knee width . 


Shoulder-hip ratio 4 



i The Medical Department of the United States Army in the World War, vol. 15, Statistics, 
Part 7, Army Anthropology, p. 234, Government Pnnting Office, Washington, 1921. 

MILES, W. R., op. at. 

1 WILDER, H. H., A laboratory manual of anthropometry. Appendix B, pp. 171 ff., Blalos- 
ton's, 1920. 

4 Biacromial diameter divided by the hip diameter between the iliac crests. 

TABLE 62. PHYSICAL MEASUREMENTS OF NINE HOMOSEXUAL YOUTHS 



Item 


Mean 


S.D. di- t. 


Item 


Mean 


S.D. di . t . 


Aee 


18.7 


0.9 


Hip diameter, tro- 






Height 


175.0 


6.18 


chanter 


32 28 


1.13 


Weight 


138.05 


15.72 


Shoulder slope 


112.33 


3.40 


Sitting height 
Sternal notch height. 
Pubic height 


88.72 
141.66 
90.78 


3.50 
5.33 
4.56 


Hip circumference 
Head circumfer- 
ence 


89.22 
56 83 


4 57 
2.11 


Shoulder diameter 


36.17 


2.63 


Knee width 


21 36 


1.05 


Hip diameter, iliac 
crest 


28.42 


2.05 


Shoulder-hip ratio. 


1.28 


.08 















their means to be somewhat smaller than those of the other male 
groups. The only statistically significant difference is between 
the mean height of the homosexuals and that of Stanford fresh- 
men. The latter probably represent a selected group so far as 



A STUDY OF MALE HOMOSEXUALS 



253 



height is concerned; on the other hand, the homosexual group 
have nearly all reached their adult height while most of the 
college freshmen have not. 

It is commonly believed that men have relatively narrower 
hips and broader shoulders than women. The shoulder-hip 
ratios of 1.34 for the college men and 1.30 for the college women 
support this belief to a slight degree only. The mean ratio 
for the homosexual group is 1.33, and there is about a fifty-fifty 
chance that the mean shoulder-hip ratio of another group of 
24 homosexuals would be as high as 1.35 or as low as 1.31. 

Because of the small number of subjects in the P.M.H. group, 
the differences between the means for this group and the other 
male groups are too small to be accepted as evidence that male 
passive homosexuals are distinguishable from the generality 
of men with regard to size or body build. 

One possibly significant finding is that the group of 24 includes 
two fairly distinct body types: the small, slender, even delicate, 
type and the large, fat, voluptuous type. The subjective 
impression of the investigator that such was the case is con- 
firmed by the following bimodal distributions of the height 
measurements: 



Height, centimeters 


Frequency 


183 to 185.9 


1 


180 to 182.9 


1 


177 to 179. 9 


5 


174 to 176. 9 





171 to 173. 9 


1 


168 to 170. 9 


2 


165 to 167. 9 


3 


162 to 164.9 


3 


159 to 161.9 


3 



Again it must be stressed, however, that the data are based on 
too few cases to justify anything beyond the most tentative of 
conclusions. 

None of the 33 cases for whom a physical examination was 
possible showed the slightest defect of the genital organs. This 



254 SEX AND PERSONALITY 

is in agreement with the findings of other investigators and with 
the statements obtained from the subjects to the effect that such 
defects are practically unknown among their groups. 

One other physical characteristic should be noted, namely, the 
amount and distribution of body hair. Many of the 33 subjects 
had but a small amount of body and leg hair and as a rule their 
beards were considerably less heavy than those of the average 
men. 1 Many of the group reported that it was necessary for 
them to shave but once every three or four days. A few cases 
were found with typically feminine distributions of pubic hair, in 
which the usual male vortex was entirely lacking. Not one of the 
group was of the extremely hairy type. These facts suggest 
the possibility of certain endocrine disturbances. It is possible, 
however, that the small group who were examined physically 
may not have been a representative sampling of the passive male 
homosexual population. Further investigation is urgently 
needed of physique in homosexuals. 

THE GROUP OF ACTIVE MALE HOMOSEXUALS 

This is the group of army prisoners previously referred to who 
were serving sentences for sodomy and had been classified by the 
prison psychiatrist as belonging to the active type. 2 The 56 
subjects tested yielded only 46 usable blanks, as several of the 
subjects were partly illiterate and omitted many items. 

The distribution of scores for this group has been shown in 
Figure 15. The mean of the distribution is +66.2, and the S.D. 
is 46.3. Both the mean and S.D. are almost identical with those 
for 130 male college students; the mean is 14 M-F points more 
masculine than that of 604 adult males in our general-population 
group. We have also seen that it is more masculine than the 

1 It may be noted that depilation, both facial and bodily, is a fairly common 
practice among homosexual male prostitutes. 

2 The authors are greatly indebted to Commanding Officer Colonel Cralle of 
Alcatraz for permission to give the M-F test to these groups and for his hearty 
cooperation throughout the study; also to Captain G. E. Hesner for his assistance 
in administering the tests and for supplying valuable case-history data for the 
group. 



A STUDY OF MALE HOMOSEXUALS 255 

mean of 42 regular army men of similar education and social 
status. 

The mean age of the group was 26.4 years; S.D. of the dis- 
tribution 3.5 years; range 17 to 46 years. These figures parallel 
closely the age data for the P.M.H. group. 

The average number of brothers for the members of the active 
group is 2.2; of sisters, 2.3. Two are "only" children. 

Half the subjects were reared in a rural, half in an urban, 
environment. 

Thirty-five were reared by both parents, two by father only, 
four by mother only, three by grandparents, and two by guardians. 

Only six of the group attended high school and only one 
graduated. A third of the remainder did not go beyond the sixth 
grade. In education, intelligence, and culture this group is 
distinctly inferior to the P.M.H. group, which makes their high 
masculine score all the more significant. 

The two favorite childhood games were baseball and football 
(no other game was mentioned more than once). Neither of 
these games was mentioned as a favorite by any of the P.M.H. 
group. 

The favorite childhood toys were bicycles and kites. In the 
P.M.H. group, kites were mentioned once and bicycles not at all. 

All of the subjects were unmarried at the time of the investiga- 
tion; a few may have been married at some time, but information 
on this point was unfortunately not secured. 

Although all had been sentenced for sodomy, only 21 admitted 
their guilt and 5 of these claimed to have been drunk at the time 
the offense was committed. The remainder denied guilt and said 
they were " framed." 

The group was classified by the psychiatrist as follows: 

1. Inferior type and inadequate personality 14 

2. Constitutional psychopathic state 13 

3. Inadequate personality and emotional instability. . 5 

4. Mental deficiency 3 

5. No mental abnormality 5 

6. Miscellaneous 6 



256 SEX AND PERSONALITY 

Further studies are urgently needed of the distinguishing char- 
acteristics of the active and passive types of male homosexuals, 
with greater care to secure groups of individuals who belong with 
certainty to one class or the other. At the same time two other 
groups should be studied: (a) homosexuals who are equally 
given to the active and passive roles, and (6) bisexuals who 
alternate between homo- and hetero-sexuality. 

GENESIS OF HOMOSEXUALITY 

Certain investigators have stressed the importance of innate 
constitution in the production of homosexual behavior, while 
others have insisted just as strongly that it is chiefly the result 
of environmental conditions. The data collected in this study 
offer interesting evidence as to the part which environment has 
played and at the same time fail to disclose any definite con- 
stitutional basis for the condition. The fact that the entire 
group of 77 passive homosexual men studied has produced only 
five children suggests that group suicide is going on. If this is 
representative of similar groups, and the homosexual population 
is dependent on direct inheritance from practicing individuals 
of this type, it could not survive for many generations. 

It is probable that active and passive homosexuality have 
different causes. With respect to the former we have no data to 
offer beyond the generally masculine M-F scores of the Alcatraz 
group. Passive male homosexuals are typically true inverts 
in the broadest sense of the term; active male homosexuals 
seem not to be. The latter are masculine in their responses, 
but the stimulus which provokes their sexual responses is a male 
instead of a female, though preferably a male of feminine per- 
sonality. It seems probable that most of them are bisexual, as 
men have been in certain historic cultures, among the Greeks, for 
example. 

As will be seen later (Chapter XV), the situation seems to be 
reversed with women. Here it is the active homosexuals who 
are the true inverts, the passives ordinarily being normally 
feminine in their attitudes and interests and perhaps usually 
bisexual. In the making of active male and passive female 



A STUDY OF MALE HOMOSEXUALS 257 

homosexuals it is probable that chance circumstances often 
play an important role, also, especially in the case of males, a 
willingness to experiment in search of new types of sexual 
gratification. The active male homosexual is rather a pervert, in 
the primary and nonmoral connotation of the term. 

SUMMARY 

1. The M-F score mean of a group of 77 passive male homo- 
sexuals aged 17 to 44 years is 27.97 with a standard deviation 
of 42 .25 . This is by far the most feminine-testing group of males 
encountered in our investigations, more feminine in fact than 
our group of outstanding college women athletes. 

2. The responses of this group tend to be feminine on all the 
exercises of the test except No. 4 (emotional response) . Exercise 5 
(interests) yields by far the most feminine scores of the seven 
exercises. 

3. It is not believed that the test responses of the P.M.H. 
group were much modified by the deliberate effort of its mem- 
bers to make themselves as feminine as possible. The subjects 
did not know the specific purpose of the test, and their scores on 
the separate exercises differ in characteristic ways from those of 
male subjects who take the test with the instruction to respond 
as they think women would do. 

4. A group of 46 male homosexuals classified as belonging to 
the active type earned a mean score slightly more masculine than 
that of unselected soldiers of comparable age and background. 
The distribution for this group was bimodal, indicating that the 
population probably contained a number of passive homosexuals. 

5. Half or more of the group of 77 passive homosexuals were 
prostitutes; practically all the 50 other occupations recorded as 
having been followed at any time were occupations which tend to 
attract effeminate men. 

6. A series of 13 physical measurements applied to 24 members 
of the P.M.H. group yielded only negative results. Several 
of the group, however, showed feminine characteristics with 
respect to quantity and distribution of body hair, 1 and appar^ 

1 See note, p. 254. 



258 SEX AND PERSONALITY 

ently there was some indication of two feminine types (small 
slender, and large voluptuous). 

7. The passive male homosexuals as a group have been 
markedly feminine in their interests and preferred activities since 
early childhood. Although exactly comparable data were not 
secured for the active group, such casual information as was 
secured presented a strikingly contrasting picture with respect 
to masculinity-femininity during childhood. 

8. Although our data offer no crucial evidence regarding the 
genesis of homosexuality in males, we believe that they lend 
considerable support to the theory of psychological conditioning. 
Additional evidence will be presented in the two chapters which 
follow. 



CHAPTER XII 

A TENTATIVE SCALE FOR THE MEASUREMENT OF 
SEXUAL INVERSION IN MALES 1 

The preceding chapter has shown that passive male homo- 
sexuals as a group rate far more feminine on the M-F test than 
any other group of men comparable in age and education. In 
fact, their responses are indicative of interests, attitudes, and 
thought trends that are essentially more feminine than mas- 
culine. We have already seen, however, that the P.M.H. group 
is not equally feminine on all the exercises of the test. Exercise 5 
(interests) contributes most heavily to their feminine rating. On 
this exercise the critical ratio of the group's divergence from our 
general population groups is 15.3. The critical ratios are also 
large on Exercise 2 (ink blots) and Exercise 7 (introversion), 
6.2 and 5.6, respectively, but these exercises are so lightly 
weighted as to contribute little to the variance of the total 
score. On Exercise 6 (personalities and opinions) the invert 
group is somewhat more feminine than the general-population 
group (C.R. 2.0), but on Exercise 1 (word association), Exercise 3 
(information), and Exercise 4 (emotional and ethical attitudes) 
their divergences from the general population of males in each 
case lacks statistical significance. When the P.M.H. means were 
computed for the four parts of Exercise 4, it was found that in 
anger and pity responses the group is near average for the general 
male population, more feminine than the latter in fear responses, 
rather more masculine in disgust responses, and markedly more 
masculine in ethical attitudes. 

It seemed desirable to investigate still more specific divergences 
of the invert group from a norm group by the tabulation of 

1 Dr. E. Lowell Kelly, assistant in the M-F investigation, is responsible for 
bringing together the factual material of this chapter. The summary here pre- 
sented is the joint work of Dr. Kelly and Dr. Terman. 

259 



260 SEX AND PERSONALITY 

responses to the individual items of the M-F test. The normal 
population selected for this purpose was a group of 98 males 
enrolled in the junior year of high school, whose item responses 
had already been tabulated for another purpose. The members 
of this group were younger than the large majority of the invert 
group and perhaps averaged a little higher in general intelligence. 
The fact that the two groups were not more evenly matched for 
age, education, intelligence, and social background detracts in 
some degree from the value of the resulting comparative data, but 
probably not to any very great extent; certainly most of the 
larger differences we have uncovered would be found to hold for 
groups more carefully matched. The choice of a norm group was 
dictated by economy, item tabulations for the high-school group 
being already available. 

The comparison of item responses was made with two purposes 
in view: (1) it would make possible the derivation of a tentative 
scale for the measurement of sexual inversion in males, and (2) 
it would afford the basis for a detailed descriptive characteriza- 
tion of the invert group as compared with noninverts of the 
same sex. 

DERIVATION OF "I" SCORE WEIGHTS 

The item comparisons could be made only for Form A of the 
M-F test, as Form B was not administered to the invert group. 
The techniques used in deriving the "I" scale were essentially the 
same as those by which response weights were derived in our 
preliminary experiments with the various parts of the M-F test, 
except that in this case two contrasting male groups were used 
instead of a male and a female group. That is, for each item 
of Form A of the M-F test the percentages of the two groups 
answering the item in a given way were computed. As each 
item has from two to four multiple-choice responses, percentages 
had to be computed for a total of some 1,560 possible responses. 
The differences between these percentages for the two groups 
were then translated into "I" score weights, a weight for each 
of the possible responses. 



MEASUREMENT OF SEXUAL INVERSION IN MALES 261 

As in the case of the M-F test, the question arose whether to 
use unit or weighted scores; in other words, whether to make the 
weights uniformly +1 or 1, according as a given percentage is 
significantly higher for one group than for the other, or instead 
to give larger weights in proportion to the magnitude and relia- 
bility of the difference. It will be recalled that in the case of the 
M-F test weighted scores were finally abandoned in favor of unit 
scores. The latter reduce the labor of scoring, are about as 
reliable as weighted scores, and possibly more defensible psycho- 
logically in that they avoid overweighting large sex differences 
that may reflect chiefly the influences of differing environments. 
The world of men is in reality a very different world from that of 
women. In the present experiment, however, we are concerned 
with two male groups, and the psychological objection to the use 
of weighted scores is less cogent. Weights have therefore been 
employed. These range from +30 to 30, the plus weights 
being arbitrarily assigned to responses more frequently checked 
by the invert group. The weights for all the responses are 
reproduced in Appendix IV, so that anyone who wishes to use 
the "I" scale will be able to provide himself with a scoring key. 
Further investigation is desirable to determine whether a nar- 
rower range of weights, say from 1 to 3, would be better than the 
wide range used. 1 

In order to ascertain whether or not the "I" scores possess a 
peculiar significance of their own, that is, whether they are 
measuring something other than that measured by the M-F 
scores proper, all the test blanks which had been filled out by the 
invert group were rescored on the basis of the "I" weights. 
The distribution of the 82 scores 2 resulting is shown in Table 63. 
As will be noted immediately, the total scores are practically all 
positive, and, on the whole, quite large. The range is almost 
two thousand points, which indicates wide individual differences 
among the members of the group and that the "I" scores are 
discriminative of these differences. 

1 The "I" score weights were computed by the formula given on p. 54. 
s ln this tabulation five cases were added to the 77 in the original P.M.H. 
group. 



262 



SEX AND PERSONALITY 



TABLE 63. FREQUENCY DISTRIBUTION OP THE "I" SCORES MADE BY 82 MEMBERS 
OF THE P.M.H. GROUP 



"I" Score 


Frequency 


+1400 to 4-H99 


2 


+ 1300 to +1399 


3 


+ 1200 to +1299 


6 


+1100 to +1199 


5 


+1000 to +1099 


5 


+ 900 to + 999 


7 


+ 800 to + 899 


8 


+ 700 to + 799 


5 


+ 600 to + 699 


6 


+ 500 to + 599 


5 


+ 400 to + 499 


1 


+ 300 to + 399 


5 


+ 200 to + 299 


1 


+ 100 to + 199 


12 


Oto + 99 


4 


- 100 to - 1 


4 


- 200 to - 101 


1 


- 300 to - 201 


1 


- 400 to - 301 


1 


Mean 


+644.0 


S.D. 


462.2 



Perhaps the most interesting feature of the distribution is its 
distinct bimodality, clearly observable in Fig. 16. By means of 
the goodness-of-fit technique it was found that such an extremely 
bimodal distribution would occur only seven times in 1,000 
as the result of chance. It appears, therefore, that the invert 
group is probably made up of two distinct personality types. It 
will be remembered that every effort was made to include only 
passive male homosexuals in this group, but in view of the 
method of obtaining the subjects it would not be surprising if a 
certain number of the active type were included. It is even 
possible that a few members of the group were not true homo- 
sexuals at all, but were merely posing as such in order to ply 
the trade of prostitution. Whatever the actual situation may 
be, the scores indicate a marked heterogeneity within the group. 

That the "I" scale is measuring something other than the 
M-F scale is shown by the fact that the correlation between the 



MEASUREMENT OF SEXUAL INVERSION IN MALES 263 

two sets of scores was found to be only .09 .08. For another 
group of subjects (46 Catholic priests in training) a correlation 
of .54 was found. Even here, however, the correlation is low 
enough to demonstrate that the "I" scores have a special 
significance of their own. The mean "I" score of this group was 
187, as compared with +644 for the P.M.H. group. The range 
for the priests in training was from 700 to +300. Relative 
to the mean M-F scores of the two groups (+12 for the student 




-1000-800-600-400-200 * 200 +400*600+800*1000*1200*1400*1600 
to to to to to to to to to to to to to to 
-801-601-401 -201 -1 M99 +399 *599 +799 *999*l 199*1399*1599*1799 
Invert Scale 

FIG. 16. Distributions of "I" scores of 82 passive male homosexuals and 
46 normal males with low M-F scores. 

priests and 19 for the homosexuals), the student priests test 
much less invert on the "I" scale. 1 

The question may well be raised as to just what a large "I" 
score signifies in terms of personality and temperament. With- 
out going beyond the empirical findings we can say that a 
person making such a score possesses a set of interests, attitudes, 
and thought trends (within the area covered by the test) approxi- 
mating those of a typical member of the invert group. To the 
extent that this group is representative of the generality of 
passive male homosexuals, it might be said that the subject in 

1 Unfortunately the group of 46 male homosexuals classified as actives (Alca- 
traz group) were given Form B of the M-F test and could therefore not be scored 
on the "I "scale. 



264 SEX AND PERSONALITY 

question resembles in personality the typical male sexual invert. 
There, however, we must stop; a man with a high "I" score may 
or may not be a practicing homosexual; for all we know he may 
have made normal heterosexual adjustments. Only extended 
and careful research can determine the extent to which a man 
whose personality deviates from the average in the homosexual 
direction is thereby predisposed to find heterosexual adjustment 
difficult. One factor which certainly limits the validity of the 
"I" scale is the fact that the invert group on which it is based 
is not truly representative of inverts in general but is composed 
chiefly of invert prostitutes. 

There can be little doubt that an "I" scale based upon large 
and strictly representative groups of homosexual and nonhomo- 
sexual subjects would be of value for research and clinical 
purposes. Subjects at the secondary school level with high "I" 
scores should be followed up in order to find out what types of 
sexual adjustment they are likely to make. If it should turn out 
that young men with such scores are in fact potential homo- 
sexuals, preventive measures might be found that would direct 
their sexual development into normal channels. 

The " I " scores may also be found useful in the legal disposition 
of the cases of homosexuality continually coining before the 
courts. Although the courts are primarily interested in the 
criminal act itself, rather than in the factors leading up to the act, 
it seems reasonable that different types of treatment should 
be accorded those found guilty of homosexual acts on the basis 
of whether they are true homosexuals or mere perverts. It 
will bear repetition, however, that the "I" score cannot be 
taken as conclusive proof of the presence or absence of homo- 
sexuality. It does constitute an entirely new line of evidence, 
which, interpreted in relation to information provided by the 
case history, assumes unique significance. In the following 
chapter, which presents 18 case histories of members of the 
P.M.H. group, the reader will find it interesting to compare the 
"I" score of each subject with the obtained behavioral evidence 
of true sexual inversion. 



MEASUREMENT OF SEXUAL INVERSION IN MALES 265 

ANALYSIS or "I" WEIGHTS OF INDIVIDUAL RESPONSES 

By noting the "I" weights assigned to the various responses 
in the manner described in the last section it is possible to 
characterize the invert group in terms of the responses which are 
most or least typical of it. A survey of the items showing 
significant differences between the two groups will afford a kind 
of composite picture of the attitudes and interests which go to 
make up the personality of the male sexual invert. 

However, in any use of the response weights for this purpose it 
is necessary to bear in mind the nature of our invert group. As 
we have already pointed out, this group is considerably older 
than the norm group with which it is compared and differs 
from the latter in experience and social background. There are 
also differences in education and intelligence, though these are 
smaller and less important for the present purpose than the 
differences in age and life experiences. Such differences, having 
little or nothing to do with sexual inversion per se, would doubt- 
less account for many of the plus and minus weights, especially 
those of small or moderate magnitudes. There is an additional 
reason why it is necessary to avoid attaching too much signifi- 
cance to the weight assigned to an individual response. The 
reliability of a single item weight is so extremely low that even 
if the groups had been more ideally matched, chance factors 
alone would have produced many weights of appreciable size. 
In the total score the chance factors tend to cancel out, as the 
weights they produce would theoretically fall as often in one 
direction as the other. Chance factors would very rarely 
produce a large "I" score, either in the plus or minus direction; 
such a score could only be obtained by the subject who makes 
a large number of responses typical (or atypical) of inverts. In 
item comparisons, however, chance factors must not be over- 
looked. In the item analysis which follows no account has been 
taken of weights below +7 or 7. Anyone who wishes to 
extend the analysis to lower weights can do so from the data 
supplied in Appendix IV. For ourselves we prefer to leave the 



266 SEX AND PERSONALITY 

more detailed comparison until data are available for larger and 
more representative groups. 

Exercise 1: Word Association. This test, as we have seen, 
does not effectively differentiate the invert group from normal 
males. Examination of the separate items shows that only 
six out of 60 have any of their four multiple responses with 
weights as large as +10 or 10. These are as follows: 

11. TRAIN engine(-lO) gown(+12) travel(+4) whistle(+2) 

16. PASS car(-3) mountain(+2) over(+2) subject(-lO) 

37. DEVIL dare(-2) evil(-l) hell(-S) tempt(+10) 

42. CELLAR basement(-7) dark(+5) furnace(+3) vegetables(+H) 

44. DESPISE coward(-6) dirt(+10) dislike(-l) flirt(+7) 

60. MACHINE engine(-l) Ford(-5) ride(+2) sew(+10) 

In No. 11, although the response "engine" has the favored 
position, "gown," a typically feminine response, was underlined 
by 18 per cent of the invert group and by not a single one of the 
norm group. This gives "gown" a weight of +12. "Engine," 
on the other hand, was underlined by only 40 per cent of the 
invert group and by 80 per cent of the norm group, which gives 
a weight of 10 for this response. 

In No. 16 the weight of 10 for the response "subject" 
probably reflects nothing more than the fact that the members 
of the norm group are still in school and are interested in passing 
their subjects. 

One might explain the weight of +10 for the response "tempt " 
as indicating a more or less subconscious feeling on the part of the 
homosexual that he is forever being tempted to engage in a 
socially disapproved type of sexual behavior. 

A plausible reason why in No. 42 the response "vegetables" is 
given oftener by the invert group and "basement" by the norm 
group is that "vegetables" represents a more housewifely 
association. A majority of the invert group have done a great 
deal of housework and some of them have held positions as cooks. 

In No. 44 the fastidious nature of the typical invert asserts 
itself in a weight of +10 for the response "dirt." It is also 
interesting to note that the invert group gives a specifically 
sexual association oftener than the norm group (+7 for "flirt"), 



MEASUREMENT OF SEXUAL INVERSION IN MALES 267 

and the norm group more often an association reflecting a 
favorable attitude toward personal courage (6 for "coward"). 
Anyone who is acquainted with male sexual inverts knows that 
physical bravery is not one of their commonly cherished ideals. 

In No. 60 the weight of -f 10 for "sew," as a response to 
"machine," undoubtedly reflects the preoccupation of the invert 
group with feminine activities. 

We append a list of the items in this exercise which carry one or 
more weights as high as 7 but less than 10. Weights in this 
range are of course less reliable than higher ones, but many of 
them strongly suggest genuine psychological differences between 
inverts and normal males. 

4. SHARP bright(+4) flat(+5) knife(-T) pin(+5) 

6. ORDER buy(-2) command(-3) neat(+7) quiet(+3) 

8. POST fence(-7) gate(+2) letter(+5) mail(+4) 

9. TENDER kind(-f 1) loving(-l) meat(-4) sore(+9) 
14. FLY airpIane(-7) bird(-H) nasty(+9) travel(+7) 
18. BOOK cover(-5) paper(-7) print (+4) read(+2) 

22. DANGER accident(-l) caution(-4) death(+7) escape(+6) 
24. FRESH cool(-7) flirt (+6) meat(-l) stale (+2) 

29. GARDEN flower(-2) fruit (+7) vegetable(-l) wceds(-4) 

30. EMBRACE arms(-6) lover(-l) mother(+3) sin(+7) 

31. HOME expenses(+7) happiness(+3) house(-6) sleep(-6) 

32. BLUSH red(-8) rose(+7) shame(-H) smile(-l) 

34. FELLOW boy(-4) friend(O) good(+7) pal(+l) 

35. CHEAT cards(-4) clerk(-f-7) crook(+l) unfair(-l) 

39. DIMPLE baby(+3) cheek(-7) hole(+5) knee(+7) 

40. KNIGHT armor(-8) brave(+4) Ivanhoe(+3) man(+l) 

41. LETTER love(+4) news(O) paper(-9) stamp(-3) 

47. SPOON fork(~2) pet(+8) silver(-l) soup(-2) 

48. CHEEK blush(-7) girl(-5) nerve(-hS) pink(-fl) 

51. SACRIFICE cards(+l) kill(-4) money(-6) mother(-f7) 

56. FAMILY brother(-7) kind(O) quarrel(-fS) sister( + l) 

57. SIXTEEN age(-8) foolish(+5) number(O) years(-H) 

Li the above list No. 4 suggests that male inverts are less 
interested than normal subjects in the knife as a tool ("knife," 
7) and more interested in music ("flat," +5) and feminine 
appliances ("pin," +5). No. 6 again reflects the fastidious 
habits of the invert ("neat/ 1 +7). In No. 8 it is not surprising 
that "post" more often suggests "letter" to the inverts and 
"fence" to the normals. One wonders whether the +9 for 



268 SEX AND PERSONALITY 

"sore," in response to "tender," has any connection with the 
frequent practice of pederasty by the inverts. In No. 14 the plus 
weight for "nasty" and the minus weight for "airplane" 
undoubtedly reflect the greater fastidiousness of the invert 
and the greater mechanical interest of the normal. In No. 22 
the weights of +7 and +6 may reflect the invert's timidity 
in the face of danger. In No. 24 it is hardly surprising that 
inverts should more often take "fresh " in the sense of "flirt " and 
normals more often in the sense of "cool." In No. 30 the plus 
weight for "sin" in response to "embrace" may reflect the 
invert's bad social conscience over the kind of embrace to which 
he is accustomed, or it may reflect merely his more intense 
preoccupation with sex. For the normal youth, "home" 
(No. 31) is a house or a place to sleep; to the invert it more often 
suggests expenses. In No. 32 the plus weight for "rose " and the 
minus weight for "red" in response to "blush" probably reflect 
the more sentimental and poetic attitude of the inverts. In 
No. 39 the minus weight for "cheek" and the plus weight for 
"knee," in response to "dimple," suggest an interesting difference 
in fetishistic associations. The +8 for "pet" in response to 
"spoon" (No. 47) is self-explanatory. In No. 48 the 7 for 
"blush" and the +8 for "nerve," in response to "cheek," may 
show the tendency to a particular slang usage in the P.M.H. 
group. The weights of 7 and +5 in No. 56 probably reflect 
the domestic conflict which so many of the invert group experi- 
enced during childhood. We feel less certain of the explanation 
for the weights in items 18, 29, 34, 35, 40, 41, and 57, though for 
most of them it would be easy to supply more or less plausible 
guesses. 

Every one of the items in Exercise 1 carries a plus weight for 
failure to respond, and in the case of 49 of the 60 the weight is 
+10 or more. The more frequent blocking of associations in the 
invert group may be a psychoneurotic symptom. 

Exercise 2 : Ink Blots. This test is even less discriminative 
between the two groups than word association. Of the 18 
items, only 4 carry a weight as high as 7 and only 2 a weight as 
high as 10. These are as follows: 



MEASUREMENT OF SEXUAL INVERSION IN MALES 269 

6. ax(-ll) boat(-4) chopper(+7) moon(-3) 

11. Indian( 5) man hanged (-7) scarecrow( 1) tassel(-f-7) 

12. chimney(-4) coil(-7) smoke(-flO) thread(+9) 
18. aman(-6) bowl(+8) cup(-5) head(-l) 

The psychological significance of most of these weights will be 
obvious. In No. 6 ax is a masculine tool associated with a degree 
of physical exertion very repugnant to the typical male invert, 
who is more often interested in such domestic appliances as food 
choppers. The domestic interest of the invert also probably 
explains the +8 weight for "bowl" in No. 18. That "cup" was 
more often checked by the norm group is probably due to the 
fact that the stimulus looks more like an athletic trophy cup than 
a household cup. In No. 12 the -7 for "coil" and the +9 for 
"thread" reflect, respectively, the mechanical interests of the 
norm group and the feminine interests of inverts. We are 
uncertain as to the meaning of the +10 weight for "smoke" 
in this item or the two weights of +7 and 7 in No. 11. 

Half of the items in Exercise 2 carry a plus weight of 10 for 
failure to respond, none bears a negative weight for omission. 

Exercise 3: Information. Although this exercise as a whole 
does not differentiate the two groups to any very considerable 
extent, there are 29 items carrying a weight of 10 or more and 
10 additional items weighted 7 to 9. Those with the larger 
weights will illustrate sufficiently the characteristic differences 
between the groups in this test. 

1. Marigold is a kind of fabric(-ll) flower(+8) grain(-ll) stone(-S) 
Inverts probably have more interest in flowers. 

2. Things cooked in grease are boiled(O) broiled(+10) fried(+6) 
roasted (6) 

The correct answer is more often given by the inverts, as would be 
expected, but so is the incorrect response "broiled." Perhaps 
one could say that the latter is a better error than "roasted," which 
is more often given by the norm group. 

3. The Yale is a kind of hammer(0) lock( 3) screen(0) wrench( 10) 
Lack of interest by inverts in things mechanical. 

5. Pongee is a kind of doth(-l) drink(+6) flower(-lO) game( 10) 

It is surprising that all the weights of any appreciable size are carried 

by incorrect responses, and that the correct answer carries a minus 



270 SEX AND PERSONALITY 

weight. One would have expected the inverts to have more knowl- 
edge of fabrics than this item indicates. 
6. The most gold is produced in Alaska(+7) New York(O) Tennes- 

see(-10)Texas(-3) 

The inverts' greater knowledge here may be due chiefly to age and 
experience. 

10. Beethoven is known as a composer(+l) painter(O) poet(+4) 
singer (10) 

More inverts know that Beethoven was not a singer, but the correct 
response received only a weight of +1. This was a surprise in 
view of the fact that so many of the inverts seemed to have musical 
interests. 

11. Most of our anthracite coal comes from Alabama( 8) Colorado( 5) 
Ohio(-ll) Pennsylvania(+18) 

The greater knowledge of industrial geography on the part of inverts 
may be due to age and experience. 

12. The number of players on a baseball team is: 7(-4) 9( 5) ll(+4) 



Inverts are notoriously little interested in games and sports. 

14. A loom is used for cooking(+3) embroidering (+3) sewing( 10) 
weaving ( 1) 

The one large weight is due to ignorance on the part of the norm 
group, but it is surprising that the correct answer was not better 
known to the inverts. 

15. Peat is used for fuel( 9) pavement (+9) plaster (+3) road-mak- 
ing^ 11) 

Peat is a relatively unknown term to the inverts. 

17. Tokyo is a city of China( 5) India(+H) Japan(+2) Russia(O) 
The one large weight is caused by the relatively greater ignorance of 

the invert group. 

18. The first American naval hero was Hull(-30) John Paul Jones(+30) 
Lawrence(+ll) Winslow(-S) 

19. Daffodils are grown from bulbs(0) cut tings ( 11) seeds(+8) 
shoots(+5) 

About equal ignorance in the two groups. 

20. The baby found in the bulrushes was Jacob (6) Jesus( 11) 
Moses(+7) Paul(O) 

Inverts appear to have more knowledge of the Bible; a good many 
of them had a religious upbringing. 

21. The boomerang is an animal( 1) plant(+10) tool(-flO) weapon( 7) 
Greater interest of the norm group in things mechanical. 

22. Minnehaha means falling leaves( 10) laughing waters(+2) running 
brooks(+3) whispering pines( 1) 

Nearly equal ignorance in the two groups. 



MEASUREMENT OF SEXUAL INVERSION IN MALES 271 

23. A correct expression is I have dove(-6) I dived(+12) He dove(- 11) 
The inverts are more fastidious in their language as in other 

things. 
25. The vessel which overcame the Merrimac was the Connecticut (+10) 

Monitor(-S) Old Ironsides(O) 
The weight of +10 for an incorrect response is out of line with the 

data for No. 18. 
28. A shilling is worth about 25 cts.(+6) 50 cts.(-9) $1.00(-4) 

$5.00(-11) 
Superior knowledge of inverts may be due to age and experience. 

37. The number of persons on a jury in the U.S. is 8(-9) 12(+8) 16(- 10) 
18(0) 

Greater knowledge of inverts about legal procedures probably due to 
age and experience and also to the fact that they live in constant 
fear of the law. 

38. The madonna is a favorite subject for music(~6) paintings(+7) 
poetry( 10) stories( 7) 

Inverts more often have artistic interests; possibly the madonna 

motif is accentuated by the mother complex. 
46. A buffet is used for books(-ll) clothes(-ll) dishes(+9) food(+7) 

Inverts have more housewifely interests. 
55. "Mennen's" is the name of cold cream( 8) perfume( 9) collar( 8) 

talcum(+16) 

Inverts are more given to the use of cosmetics. 
58. "Charades" is a running game( 9) game of chance(+3) guessing 

game(+6) kissing game( 11) 
Charades is a feminine indoor game that probably holds more interest 

for inverts than for the norm group. 
61. Babies should be weaned at about 3 mos.( 13) 6 mos.(-lO) 12 

mos.(+7)2yrs.(+13) 

Inverts display more "maternal" or infantile attitudes; are more 
interested in nursing. 

65. A birthright was sold for a mess of pottage by Cain( 6) Esau( + 12) 
Isaac(-7) Judas(-2) 

Inverts have more knowledge of the Bible and probably greater 
interest in religion. 

66. Beam scales illustrate the principle of buoyancy(+l) elasticity (+10) 
leverage( 2) magnetism(+l) 

The only considerable weight is for a wrong response by the inverts 

to an item concerning mechanics. 
69. "Nevermore" was spoken by a general (11) parrot (+1) raven(+14) 

woman (4) 
Inverts are probably more interested in poetry, or at least in poetry 

of a certain kind. 



272 SEX AND PERSONALITY 

70. A famous portrait painter was Rosa Bonheur( 8) Mozart (2) 

Reynolds(+10) Rubens(+l) 
Inverts are more interested in art. 

Of the 70 items in this exercise, 11 carry plus weights of 10 or 
more for omission and two minus weights of this magnitude. 
The difference could mean either that the invert group has less 
information than the norm group or that the latter is less timid 
about guessing. On the other hand it may reflect nothing more 
than the school youth's greater willingness and readiness to 
respond to pencil-and-paper tasks. 

All in all, the group differences in this exercise seem to us less 
significant psychologically than those in Exercise 1 ; the influence 
of age and experience often seem to outweigh personality factors 
in producing sizable weights. 

Exercise 4: Emotional and Ethical Attitudes. This is another 
test which differentiates to only a slight extent between the two 
groups. The several parts of the exercise will be treated sepa- 
rately for weights of 7 or higher. 

a. Anger Responses. 
7. Being snubbed by an inferior VM(+S) M (-2) L(+2) tf (-4) 

That inverts more often respond to this situation with "very much" 
anger probably reflects their sensitiveness and introvertive tend- 
encies. 

11. Seeing boys make fun of old people VM(-7) M(+4) L(+3) N(+ 10) 
The explanation is not obvious to us. 

12. Seeing an honest official thrown out of office by politicians VM( 4) 
M (-2) L(+3) N(+7) 

Interpretation uncertain; possibly the inverts are less interested 
in civic affairs, or possibly they regard the phenomenon as too 
common to get excited about. 

13. Seeing a person laugh at a cripple VM(-3) M(+3) L(+4) N(-IO) 
We see no obvious explanation of the fact that both extremes of 

response should be more frequent with the norm group. 

16. Seeing someone trying to discredit you with your employer VM(Q) 
M(0) L(-l) N(+7) 

Interpretation of weight for N not clear. Perhaps this represents a 
common experience for the inverts. 

17. Seeing someone laugh when a blind man runs into an obstacle VM(Q) 

L(+3) N(-1Q) 



MEASUREMENT OF SEXUAL INVERSION IN MALES 273 

The N weight here is in line with that in No. 13, but not with the 
N weight in No. 11. Reason not obvious. 

For the 17 items of the test the total of all the plus weights 
assigned to VM and M is 46; the total of the minus weights is 
36. The total of the plus weights assigned to L and N is 65 ; the 
total of the minus weights is 54. The difference between the plus 
and minus totals, though small, is in each case in the direction 
of a larger plus total. It seems at first a contradiction that the 
inverts should score higher both for VM and M on the one hand 
and for L and N on the other. The contradiction disappears 
when we cast up the totals of VM , M , and L for comparison with 
the total of N taken alone. These are, for VM, M, and Z, +77 
and 48; for N, +34 and 42. The responses accordingly 
indicate slightly greater proneness of the invert group to say that 
they would experience at least some degree of anger in situations 
of the kind presented. 

b. Fear Responses. 

2. Being lost VM(-l) M(-7) Z,(-4) #(+7) 
Age and experience of inverts may account for the +7 weight for N. 

4. Becoming deaf or blind VM(-7) M(-7) L(+6) N(+W) 

The inverts are probably more careless regarding physical misfortunes. 

5. Bulls VM (+9) If (+2) Z,(-3) #(-4) 

Inverts are characterized by greater physical timidity. 
11. Garter snakes VM(+7) M(+l) L(+3) N(-S) 

The inverts here show a feminine type of timidity. 
16. Negroes VM(+W) M(+6) L(+7) N(-9) 

We do not know why the inverts stand so much more in fear of 
negroes than do the norm group, unless it be that negroes take 
advantage of their feminine timidity to bully them. 
18. Punishment in the next world VM(+l) Jf(-4) Z,(+8) #(-5) 
Explanation of L weight not clear, especially as that for N is in the 

opposite direction. 

20. Windstorms VM(+10) N(+5) Z,(-2) #(-1) 
Physical timidity probably an adequate explanation. 

The plus weight total for VM and M is +107; the minus 
weight total, 47. The corresponding totals for L and N are 
+58 and 69. In other words, the invert group inclines more 
to "much" or "very much" fear, the norm group to "little" or 



274 SEX AND PERSONALITY 

"none." The totals for VM , M, and L are +142 and -67; for 
N taken alone, +23 and 49. There can be little doubt that 
these figures express a genuine difference between the two groups 
in their attitudes toward situations requiring physical courage. 

In this section omission receives only plus or zero weights, for 
12 of the 20 items the weight being as high as 10. 

c. Disgust Responses. 

2. A butcher shop VM (+4) M (+3) L(+7) #(-8) 
The inverts are a little more fastidious. 

12. Seeing a woman smoking VM (-9) Af(-4) L(+2) N(+6) 
Difference probably due to age and experience. 

15. Soiled or ragged finger nails VM(+2) M(+l) L(-M) #(-7) 
Again the inverts are slightly more fastidious. 

From the small number of items with a weight of 7 or higher 
one would conclude that no appreciable difference exists between 
the two groups in disgust response. However, an interesting 
difference is disclosed when we compare weight totals. For VM 
and M the plus weights for the 18 items total only +31; the 
minus weights, 63. For L and N the totals are +64 and 
37. It appears, therefore, that the inverts are decidedly less 
prone to experience disgust than the norm group, at least in 
situations of the kind with which we are here concerned. This 
may possibly be due to the age difference, but we are inclined 
to think that it reflects a dulling of the invert's sense of disgust 
by the perverted sexual practices to which he is accustomed. In 
view of the nature of these practices it is surprising that the 
inverts are as capable of disgust as the test indicates them to be. 

All but one of the weights for omission are zero or plus, nine of 
them being as high as 10. 

d. Pity Responses. 

1. A bee that is drowning VM (-5) J(-7) L(+l) tf (+5) 

7. Overworked horses FJf(-3) Jlf(-l) (+8) tf(+4) 

8. Overworked children VM(-V) M(0) L(+2) N(-W) 

9. A fly caught on sticky fly paper VM(+9) M (-3) L(+l) #(-3) 
10. An underfed child VM(Q) M(+2) I(-l) #(-10) 

13. A baby bird whose mother is dead 7Jf (-5) M(-5) (+8) #(+7) 

14. A wounded soldier who must beg for a living 7M(-8) M (+4) L(+7) 
N(+9) 



MEASUREMENT OF SEXUAL INVERSION IN MALES 275 



The situations presented in items 1, 7, 13, and 14 arouse less 
pity in the inverts than in the norm group. The opposite is the 
case for the situations presented in items 8, 9, and 10. Probably 
the fact that high-school youths have not often come in contact 
with overworked or underfed children accounts for the weights 
in items 8 and 10. Why inverts should feel so much more pity 
than the norm group for the fly caught on sticky fly paper is 
puzzling, especially as they feel less pity than the norm group 
for the drowning bee. 

The totals of the VM and M weights for the 15 items are +25 
and 53; for the L and N weights, +63 and 37. On 
the whole, therefore, the inverts seem to be less pitying 
than the norm group, but the difference may possibly be due 
to age. 

Weights for omission are about equally divided between plus 
and minus; for four items it is 10 and for three +10. 

e. Ethical Judgment. It will be recalled that this test requires 
subjects to draw a circle around 3, 2, 1, or to indicate whether 
the act mentioned is "extremely wicked," "decidedly bad," 
"somewhat bad/' or "not really bad." Of the 28 items in the 
test, 15 received a weight of 7 or more. 



M-F test items 


3 


2 


1 





1. Picking flowers in a public park 


-10 


- 8 


-1-1 


+11 


2. Stealing a ride on a truck 
3. Telling a lie to avoid punishment . 
6. Making fun of cripples 


- 9 
-15 

- 3 


- 7 
- 3 
+ 1 


-3 

+4 
4-3 


+ 9 
4-17 
4-10 


7. Usinc slanc . . . 





- 5 


4 


+ 7 


9 Boys smoking before they are 21 


6 


7 


+5 


4- 4 


1 1 . Moderate drinking 


- 5 


11 


+1 


4-10 


12. Excessive drinking 


- 3 


+ 1 


+2 


4-10 


14. Swiping fruit out of orchards 


- 8 


- 8 


+6 


+ 7 


15. Laziness 


- 3 


- 5 


+ 1 


4-10 


16. Going to bed without saying your prayers . . . 
18. Boys fighting. 


^ 



- 4 
- 8 


-3 



4- 7 
+ 5 


19. Being a slacker in time of war. . . . 
23. Not standing when " Star Spangled Banner" is 
nlaved 


-11 
- 7 


+ 5 
- 1 


+9 
+4 


4- 8 
4- 8 


27. Having fits of temper 


- 4 


- 6 


4-3 


+12 













276 SEX AND PERSONALITY 

There is little to be said about the above items beyond the fact 
that all of them show the inverts less severe than the norm group 
in their ethical judgments. With one exception the groups differ 
rather uniformly from item to item; telling a lie to avoid punish- 
ment is conspicuous for the magnitude of the difference. Of the 
entire 28 items in the test there were only two in which received 
a minus weight, and only two in which 3 received a plus weight. 
The totals of the plus and minus weights for 3 and 2 were +18 
and 229; of the weights for 1 and 0, +243 and 24. Age 
doubtless accounts for a considerable part of this tremendous 
difference, but the low associations of the inverts and their 
distorted outlook on life are probably the chief factors. All of 
them are habitual offenders against the law and many of them 
have had jail or prison experience. 

/. Preferences. Four of the 7 items in this list have a weight of 
7 or higher. 'They are as follows, the response "1" indicating 
preference for the first thing mentioned in each item, "2 " prefer- 
ence for the second, and "S" same liking for the two. 



M-F test items 


1 


2 


S 


1. (1) Make plans (2) Carry out plans. . 


+ 9 


- 8 


-1 


2. (1) Work involving many details (2) Work involving 
few details 


- 9 


+ 10 


-1 


5 (1) Work with men (2) Work with women . . . 


11 


-f 11 


+3 


7. (1) Live in the country (2) Live in the city .... 


- 8 


+ 8 


-2 



The greater preference of the inverts for making plans, as con- 
trasted with carrying out plans, is probably a reflection of their 
extreme laziness. This may also explain the greater preference 
of this group for work involving few details. That passive male 
homosexuals usually prefer to work with women rather than 
with men is well known; we have commented in the preceding 
chapter on the feminine character of most of the occupations 
which the members of this group have followed. The preference 
of the invert group for city life needs no comment. 

Exercise 5 : Interests. This exercise, as we have seen, is the 
one which differentiates between the two groups more effectively 
than any other in the M-F test. 



MEASUREMENT OF SEXUAL INVERSION IN MALES 277 



a. Occupational Preferences. "Like" has a plus weight of 7 or 
more for the following occupations out of the total list of 25: 
private secretary (+30), singer (+24), nurse (+24), dressmaker 
(+22), florist (+13), music teacher (+13), chef or cook (+10), 
social worker (+10), novelist (+9), artist (+8), librarian 
(+7), and journalist (+7). All these occupations have negative 
weights for "dislike " ranging from 8 to 25. The occupations 
with negative weights for "like" are: forest ranger (12), 
soldier (11), building contractor (9), draftsman (8), and 
auto racer (7). These have for "dislike" positive weights 
ranging from +5 to +9. All these data are in line with what is 
known about the occupational preferences of male inverts. 

b. Preferences for People. Of the 12 types of people listed, five 
have a weight of 7 or more. These are as follows: 



M-F test items 


Like 


Dislike 


Neither 


8. People who spend freely. . 
9. People with gold teeth 
10. Tall women 


+ 19 

- 8 
+ 6 


-14 
- 2 
- 8 


-7 

+3 
+9 


11. Men who take the lead. . . 
12. Mannish women 


- 9 
+ 10 


+ 6 
- 9 


+6 
-3 











The greater preference of this group of inverts for people who 
spend freely needs no comment when it is remembered that most 
of them are prostitutes. That they less often express a liking 
for people with gold teeth probably means that they are better 
acquainted than the norm group with dentistry styles, certainly 
a feminine trait. Their greater liking for tall and mannish 
women is a common preference of male inverts. They dislike 
men who take the lead because leadership is a masculine 
trait. 

c. Reading and Movie Preferences. Six of the 14 items in this 
list have a weight of 7 or more. The inverts are characterized 
by greater liking for movie love scenes, poetry, and dramatics, 
and by greater dislike of adventure stories, mechanics, and 
science. Comment is hardly necessary. 



278 



SEX AND PERSONALITY 



M-F test items 


Like 


Dislike 


Neither 


3. Movie love scenes 


+13 


-10 


- 6 


4. Poetry 


+ 5 


- 7 


+ 1 


7. Adventure stories 


-10 


+ 5 


+11 


9. Radio magazines 


-14 


+ 9 


+ 3 


10. Chemistry 


- 4 


+ 8 


- 4 


11. Dramatics 


+ 13 


-11 


- 6 



d. Activities and Foods. Here we find 12 of the 20 items with a 
weight of 7 or more. Those with plus weights for "like" are: 
charades (+11), collecting flowers (+8), hopscotch (+7), and 
cooking (+7), all of which rate high for femininity. The items 
with a negative weight for "like" are: hunting (30), horseback 
riding (10), repairing a door latch (8), and strict Sunday 
laws (8). Here as elsewhere the inverts tend to repudiate 
whatever is distinctly masculine. 

e. Literary Preferences. Seven of the 23 items in this list have 
a weight of 7 or higher. The first 5 listed below illustrate the 
feminine sentimentality of the male invert; the last 2 his dislike 
of anything that savors of sternness or adventure. 



M-F test items 


Like 


Dislike 


Neither 


2. Lorna Doone, by Richard D. Black more 
3. Through the Looking Glass, by Lewis Carroll. . . 
5. Daddy Long Legs, by Jean Webster 


+ 7 
+ 8 
+ 10 


- 7 
-11 
-4 


-3 
-4 
-3 


11. Rebecca of Sunnybrook Farm, by Kate Douglas 
Wieein 


+ 10 


- 2 


-3 


12. Christmas Carol, by Charles Dickens 
13. The Man without a Country, by Edward Everett 
Hale 


+ 9 
- 8 


- 3 
+ 11 


-2 
-1 


20. Biography of a Grizzly, by Ernest Se ton-Thomp- 
son . 


- 9 


+ 4 


+5 



/. Preferences for Drawing. Only 2 of the 8 items have a 
weight as high as 7. 



M-F test items 


Like 


Dislike 


Neither 


2. Children 


+ 10 


-7 


-4 


6. Flowers 


+ 8 


-6 


-5 



MEASUREMENT OF SEXUAL INVERSION IN MALES 279 



g. Preferences for Newspaper Reporting. 



M-F test items 


Like 


Dislike 


Neither 


2. Sporting news 
3. Musical events 
4. Theatrical news 


-23 

+ 5 
+ 12 


+ 12 
-11 
-12 


+ 8 
- 1 
-10 


5. News oddities. . . 


+ 7 


- 9 


A 



k. Travel Preferences. 



M-F test items 


Like 


Dislike 


Neither 


2. Hunt lions in Africa. . 


-28 


+ 15 


+2 


5. Visit many famous battlegrounds. 
6. Visit many manufacturing plants. 
1 1 . Learn about various religions 


- 8 
- 7 

+ 3 


+ 9 

+ 5 
- 7 


+3 
+ 1 
+2 



The data in the last three sections can be summarized by 
saying that the inverts have more interest than the norm group 
in children, flowers, musical events, theatrical news, news 
oddities, and religion, less interest in war and manufacturing, and 
very much less interest in sporting news or such dangerous 
adventure as lion hunting. 

In Exercise 5 there are 19 weights of +10 or more for omission, 
and 8 of 10 or more. 

Exercise 6 : Personalities and Opinions. The first half of this 
exercise, dealing with preferences for historical and contemporary 
characters, logically belongs with Exercise S. Twelve of the 28 
characters listed have weights of 7 to 20. 



M-F test items 


Like 


Dislike 


Neither 


2. P. T. Barnum. . . . 
7. Cleopatra 


+ 9 
+19 


- 7 
-10 


- 6 
- 8 


20. Lenin . . 


+ 7 


+ 5 


- 1? 


22. Aimee McPherson 


+13 


- 8 





24. Florence Nightingale 


+ 7 


- 3 


- 4 


4. Daniel Boone. ... 
6. Kit Carson 


-14 

-17 


+ 6 
+ 8 


+11 
+ 7 


13 Congressman Volstead 


- 9 


+11 


- 2 


14. Booker T. Washington . . .. 
IS Ulysses S Grant 


- S 
-20 


+ 8 
+11 



+11 


17 Herbert Hoover 


2 


+ 8 


- 1 


26 Billy Sunday 





+ 9 


- 3 











280 SEX AND PERSONALITY 

Perhaps the only surprises in the above data are the greater 
preference of the inverts for P. T. Barnum and Lenin. Perhaps 
the radicalism of the inverts may account for the weight of +7 
for Lenin, and their theatrical interest for the +9 weight for 
Barnum; certainly neither character could be classed as belonging 
to the feminine type. Nothing could better characterize the 
typical male invert than the weights of +19 for Cleopatra and 
20 for Ulysses S. Grant. 

In the second part of Exercise 6, in which the subject is asked 
to judge the truth or falsity of common opinions, only one of the 
14 items has a weight as high as 7. The inverts more often 
regard as false the statement that Lincoln was greater than 
Washington, but the weights are not large: true (5), false (+7). 
One would rather have expected the preference to fall in the 
opposite direction. This section of the exercise is interesting 
in the fact that it shows the inverts do not differ from the normals 
in the more strictly intellectual attitudes. 

In Exercise 6 three items have weights of +10 or more for 
omission, and eight of 10 or more. 

Exercise 7: Introvertive Response. Although this exercise 
does not differentiate very effectively between the two groups, 
half the 42 items have a weight as high as 7. They are as 
shown in the table on page 281. 

Three of the items (4, 8, and 34) reflect the fastidiousness 
of the invert group, the others their introvertive and psycho- 
neurotic tendencies. Many of them seem to have been more 
or less maladjusted from childhood, judging from the frequency 
with which they have had imaginary companions (2) and 
disciplinary trouble at school (28). Their greater fear of the 
dark (18) and their night terrors (38) are also suggestive of a 
neurotic childhood. Other items reveal them as belonging to 
the shut-in type of personality (1, 15, 40), as nervous and subject 
to anxiety (12, 17, 41), lacking self-confidence and physical 
courage (10, 32), weak in self-control (7, 31), indolent and hypo- 
chondriacal (9, 21, 35). Item 17 (feeling about to "go to 
pieces") has by far the heaviest weights in the list: +30 and 
30. Next is item 12 (worry over possible misfortunes) with 



MEASUREMENT OF SEXUAL INVERSION IN MALES 281 



M-F test items 



Yes 



No 



1. Do you like most people you know? 

2 Did you ever have imaginary companions? 

4. Do you rather dislike to take your bath? 

7. Do you feel yourself to be lacking in self-control? 

8. Are you extremely careful about your manner of dress ? . 

9. Do you work mostly by fits and starts? 

10. Do you shrink from facing a crisis or difficulty? 

12. Do you worry much over possible misfortunes? 

15. Have you ever kept a diary? 

17. Do you ever feel that you are about to "go to pieces"?. 

18. Are you often afraid of the dark? 

21. Do you usually enjoy your meals? 

26. Do you ever have the same dream over and over?. . . . 
28. Were you ever expelled from school, or nearly expelled? 

31. Does it make you angry for people to hurry you? 

32. Can you stand as much pain as others can? 

34. Would you like to wear expensive clothes? 

35. Do you feel tired a good deal of the time? 

38. Are you often frightened in the middle of the night?. 

40. Can you do good work while people are looking at you? 

41. Do you feel like jumping off when you are on a high 
place? 



-10 
+ 10 

- 6 
+ 9 
+ 7 
+ 10 
+ 8 
+ 19 
+ 13 
+ 30 
+ 7 

- 1 
+ 7 
+ 7 
+ 7 

- 6 
+ 12 
+ 10 
+ 7 

- 5 

+ 11 



+ 11 

- 9 
+ 7 

- 8 

- 7 
-10 

- 9 
-19 
-12 
-30 

- 6 
+ 11 

- 6 

- 5 

- 6 
+ 7 
-12 
-10 

- 6 
+ 7 

o, 



weights of +19 and 19. One must remember that the mem- 
bers of the invert group have good reason to feel themselves 
social outcasts; many of them have suffered legal punishment 
for their homosexual practices and most of the others live in 
perpetual fear of the law. Their psychoneurotic symptoms are 
no doubt the joint product of their adult experiences and their 
early conditioning, with perhaps some contribution from heredity 
in individual cases. 

Omitting items 4, 8, 11, 34, and 42 (all of which bear on 
fastidiousness), we have for the remaining 37 items totaled the 
plus and minus weights carried by the introvertive, maladjusted, 
or psychoneurotic responses. The first total (unfavorable to 
the inverts) is +252; the second (unfavorable to the norm 
group) is 9. Only four introvertive responses carried minus 



282 SEX AND PERSONALITY 

weights and none of these was greater than 3. The four 
exceptional items are 3, 14, 16, and 39. 

Failure to respond is weighted +10 or more for ten items, and 
10 or more for four items. 

SUMMARY 

1. On the basis of item tabulations for the passive homo- 
sexuals and a norm group of 98 males, "I" (invert) score weights 
have been derived for all the possible responses in Form A 
of the M-F test; this permits the computation of a total "I" 
score for any individual subject. 

2. The "I" scores so derived have only a moderate correlation 
with total M-F scores. 

3. The distribution of "I" scores for the invert group covers 
a wide range and has two modes, indicating that the group 
probably contains a considerable number of active homosexuals. 

4. The 1560 "I" score weights of Form A were examined and 
the more reliable ones utilized in deriving a composite picture 
of the invert personality. Some of the leading features of this 
composite picture are the following: 

a. The interests, attitudes, and thought trends of this invert 
group are more or less typically feminine throughout the test. 
In general this tendency is evident in all the exercises of the test, 
though in some more than in others. 

b. The feminine personality of the inverts appears in their 
fastidiousness with respect to dress, cleanliness, and care of 
person; in their preoccupation with domestic affairs; in their 
preference for feminine types of occupations and for working 
with women rather than with men; in their fondness for senti- 
mental movies and romantic literature; in their feminine timidity 
when faced by physical danger; in their religious interests; and 
in their liking for literature, art, music, and dramatics. 

c. The feminine personality of the inverts is evidenced nega- 
tively by their repudiation of everything that is characteristically 
masculine: aggressive leadership, energetic activity, physical 
courage, masculine pursuits, and interest in warfare, adventure, 
outdoor sports, science, and things of a mechanical nature. 



MEASUREMENT OF SEXUAL INVERSION IN MALES 283 

d. The inverts show evidence of an excessive amount of sex 
consciousness, especially consciousness of the forbidden nature 
of their sex lives. 

e. In their anger and pity responses the inverts differ but little 
from the norm group, but their fear responses are stronger and 
their disgust responses weaker. 

/. One of the most marked differences between the two groups 
appears in the lax ethical standards of the inverts; whether this 
difference would be found for a strictly random sampling of the 
entire invert population cannot be determined without further 
data. 

g. The introvertive and psychoneurotic tendencies of the 
invert group are reflected in a great variety of responses which 
indicate social maladjustment, nervousness, lack of self-confi- 
dence, a low degree of self-control, and a marked tendency to 
worry and anxiety. 

5. The various parts of the M-F test agree closely with respect 
to the personality traits which distinguish the inverts from the 
norm group, each exercise confirms the others, and the resulting 
picture is surprisingly harmonious. Further, the characteriza- 
tions of the group based upon analysis of the item responses are 
reinforced by the case-history data presented in the chapter 
which follows. The interests and attitudes of inverts as dis- 
closed by the M-F responses can hardly be explained on the 
theory that they are superficial affectations; we are convinced 
that on the whole they give an essentially correct picture of 
the actual type of personality most prevalent in this group. 



CHAPTER XIII 
CASE STUDIES OF HOMOSEXUAL MALES 

This chapter will present the first of three series of case studies 
of homosexual and other subjects. We recognize that although 
case studies are of great value to the judicious and cautious 
reader, they are likely to lure the incautious into unwarranted 
generalizations. It is therefore desirable to repeat that the 
behavioral correlates of high and low M-F scores have been little 
explored and that the reader should constantly be on guard 
against the temptation to assume that behavior exemplified by a 
particular case is necessarily typical. With this warning we 
shall present first the case studies by Dr. E. Lowell Kelly of 
18 subjects of the group of passive male homosexuals. 

One further warning should be given. In the case of the 
P.M.H. group it was necessary to depend upon the testimony 
of the subjects themselves for most of the factual information 
of the case histories, and in most instances there was no possi- 
bility of checking the accuracy of their reports. It is not at all 
unlikely that homosexual subjects, even when trying to give a 
truthful account of their early experiences, are prone to exag- 
gerate certain experiences and to forget others. For example, it 
would not be surprising if they tended to exaggerate with respect 
to the early age at which their invert tendencies appeared, for 
practically every member of the P.M.H. group believes that his 
condition is inborn. He prefers to think of his personality not 
as something malformed by circumstance, but as a product of 
nature. It is for him therefore no less "natural" than the 
personalities of others and one for which he could have no 
individual responsibility. However, even when we allow for a 
certain amount of error from this source, the case studies here 
summarized are, we believe, extremely enlightening. The 
frequency with which certain factors appear offers impressive 
evidence with respect to the environmental causes of sexual 
inversion. 

284 



CASE STUDIES OF HOMOSEXUAL MALES 285 

The information for the case histories of this chapter was all 
obtained by Dr. Kelly in personal interviews with the subjects. 
Each interview followed a definite routine outlined in the case- 
history record blank which is reproduced in Appendix IV. As 
examination of this blank will show, an attempt was made to 
secure comparable information for all the subjects on physical 
appearance, personality traits of each parent, treatment of child 
by each parent and resulting attitude of child, sibling relation- 
ships, relationships with other children, favorite toys, games, and 
other activities, educational history, school subjects liked and 
disliked, social life after childhood, religious interests, ideals, 
occupational history, and sex experiences, including autoerotic 
practices and both homosexual and heterosexual relationships. 

We regard such a routinized procedure as essential for securing 
comparable data, but it has the disadvantage of giving a some- 
what stereotyped character to the case histories. In the sum- 
maries which follow 1 the effect of stereotypy has been to some 
extent mitigated by omitting from each history a considerable 
amount of informational detail that seemed to have no explana- 
tory value. 

The 18 P.M.H. cases are presented in order of femininity of 
score, beginning with the most feminine. Both the M-F scores 
and the "I" (invert) scores are given for the separate exercises as 
well as for their total. In reading the "I" scores it should be 
kept in mind that plus scores are in the direction of inversion. 
The mean "I" score for 77 male homosexuals classified as 
passive is +644, and the S.D. of the "I" score distribution is 
462. The series of this chapter will be designated by the letter 
H with numerical subscripts to indicate individual subjects. 

H-l. "Lady M" 2 

Exercise 1234567 Total 

M-F score -4 0-29-25 -30 - 15 - 2 -105 

"I" score -50 -6 +140 +27 -9 +154 +133 +389 

1 The summaries were prepared by Dr. Kelly and the senior author from the 
original case data. 

' The names given here are fictitious. However, they are typical of the group 
discribed. 



286 SEX AND PERSONALITY 

The M-F total is 3.4 standard scores more feminine than the 
mean of third-year high-school boys or than the mean of any of 
our female groups. The "I" score is near the 30th percentile 
of the P.M.H. group for inversion. The M-F score for Exercise 4 
(emotional and ethical response) is feminine, which is unusual 
for male homosexuals. 

H-l, age 26, is in a state prison serving a sentence for fellatio. 
Son of immigrant parents. Father now about SO years old, 
mother 54. Father cold, domineering, ill-tempered and much 
feared by his children; is a confirmed alcoholic. Mother nervous 
and emotional. 

Subject is below average in height and weight and has never 
been very strong. Voice soft and gait effeminate. 

Was the first of seven children, having two brothers and four 
sisters. His mother has repeatedly told him that she wanted a 
girl at the time of his birth. She was very fond of the child, 
however, and decidedly demonstrative in her affection toward 
him. This, together with the cold and domineering attitude 
of the father, resulted in the mother becoming the favorite 
parent. The relationship between the mother and son has 
always been very intimate. The feelings toward the parents 
soon became generalized. As a child the subject was unable to 
understand men but felt quite at ease in the company of women 
and girls. In spite of ample opportunity for association with 
boys he spent most of his time with girls, playing with them 
almost exclusively. He crocheted, knitted, and took part in 
their games. He was embarrassed in the presence of other boys 
and would not go swimming with them because he was ashamed 
to undress in their presence. 

H-l had very little formal education, not more than three years 
in all. His favorite studies were reading, history, and geography ; 
arithmetic was much disliked. He is fond of music and plays the 
violin "by ear." 

About the only type of work followed has been in hotels as 
bellboy, bus boy, or waiter, although on two or three occasions 
he has gone to work in lumber camps, resolutely determined to 
become more masculine. At the present time, however, he is 



CASE STUDIES OF HOMOSEXUAL MALES 287 

still effeminate and constantly refers to himself as "she." He 
was dressed as a boy throughout childhood, but liked to masquer- 
ade as a girl whenever the opportunity presented itself. 

When still very young he was greatly thrilled at the sight of 
strong men, and he practically always falls in love with men of 
sturdy, athletic build. He has never had a heterosexual experi- 
ence. Although he is fond of women companions, they have no 
sexual attraction for him. 

He first engaged in sodomy at the age of eighteen with a male 
lover. At the present time he practices both fellatio and sodomy, 
always playing the passive role. He prefers fellatio, however, 
and was arrested and imprisoned charged with this crime. 

The subject is convinced that his homosexual tendencies are 
inborn. 

Significant facts: firstborn child; domineering and ill-tempered 
father; overaffectionate mother; lack of stimulus to play with 
boys. 

H-2. "Louise" 

Exercise 1234567 Total 

M-F score -6 -2+2-34-40-8 -3-91 

"I" score -62 +20 +150 -151 +161 +42 +31 +191 

The M-F total is more than 3 S.D. below the mean of high- 
school boys, and more feminine than the mean of all our female 
groups except domestic employees and mothers of gifted children. 
Exercise 4 is also feminine, contrary to the rule for this group. 
The " I " score is one S.D. less invert than the mean of the P.M.H. 
group. 

H-2, age 34, is a female impersonator in dancing and acrobatic 
stunts for a small eastern circus. He is small and extremely 
effeminate in both speech and behavior. Seems to be neuras- 
thenic; is nervous, shy, and apparently unstable. 

H-2 was the first of 12 children, 8 of whom are still living. 
The father, now 52, is described as brutal and autocratic in his 
attitude toward the child. Between the two there was no 
companionship or intimacy. The mother, now 51, was always 
the favorite parent. Subject spent nearly all his time with her 
and the relation between them was unusually close. Mother 



288 SEX AND PERSONALITY 

consistently catered to his every wish and desire. Subject was 
also the favorite of his maternal grandmother, who lived with 
the family. 

As a child H-2 played more with boys than with girls, although 
the boys usually regarded him as a sissy. His favorite activities 
were ball, books, chess, and croquet. His favorite books were 
romances and he much disliked "Wild West" stories. He 
graduated from high school but did not attend college because of 
lack of funds. Most liked subjects, grammar and algebra; most 
disliked, botany and civics. 

Is decidedly social; attends shows, parties, and dances, pref- 
erably in the company of men. His friends are usually older 
than himself and are chosen largely for their beauty. Member 
of United Brethren Church, but is not very religious now. His 
ideal (public) character is Woodrow Wilson; the finest person 
he has ever known is one of his girl cousins. 

H-2 has spent the last eight years as a circus performer; had 
previously worked six years in a paper mill and 18 months as a 
bus boy. He had also served two years in the U.S. Army in a 
camp hospital. 

Subject claims to have had no conscious sex experiences until 
the age of 18. At that time he learned to masturbate and 
thereafter practiced the act regularly. He believes that this 
made him more effeminate. His first homosexual experience 
occurred in the same year, when he played the passive role of 
sodomy with a first cousin. He first practiced fellatio about 
six months later with an office boy with whom he was in love. 
Both practices have been continued and both have been com- 
mercialized whenever it was necessary to make a little "extra 
money." The chief sex attraction of men for the subject is said 
to be an olfactory one. 

H-2 asserts that he is attracted to women as well as to men. 
On one occasion he was engaged to be married, but he broke the 
engagement because he was afraid he could not give up his male 
lovers. He receives some gratification from heterosexual 
experiences and visits prostitutes whenever necessary to keep up 
appearances with his friends. 



CASE STUDIES OF HOMOSEXUAL MALES 289 

Some of the case-history conferences with this subject were 
held at a time when he was "in love" with a young musician. 
He was completely infatuated with the lover and thought of him 
constantly. The least act of indifference on the part of the 
object of his affections was sufficient to induce grief almost to 
the extent of despair. His general attitude was typically that 
of a young girl infatuated with a man who cares little or nothing 
for her. 

Subject worries a great deal over his condition and would like 
to become normally sexed. However, he feels that the affliction 
is inborn and incurable. 

Significant facts: antagonism toward father; favorite of mother 
and grandmother; played more with boys than with girls; no 
tendency to transvestism; is bisexual, but with stronger leaning 
toward homosexuality; M-F score extraordinarily feminine, even 
on Exercise 4; low "I" score. 

H-3. "Adela" 

Exercise 1234567 Total 

M-F score -1 -2-21 +19 -68 -7 -3-83 

"I" score +7 +13 +147 +40 +340 +5 -20 +532 

The M-F score approximates the median for high-school girls 
and is 3 S.D. less masculine than the mean of high-school boys. 
Exercise 4 (interests) is especially feminine. The "I" score is 
slightly less invert than the mean of the P.M.H. group. 

H-3, age 27, was the youngest of three brothers, and the most 
favored by his parents. The father, who died when the son 
was about 12 years old, is described as affectionate in his treat- 
ment of the boy but somewhat moody. The mother, a trained 
nurse, died seven years ago. She is described as emotionally 
unstable and very demonstrative in her affections toward the 
boy. A close intimacy developed between them and the son 
worshiped her. Both parents were Catholics. 

After the death of his mother H-3 lived with an uncle and aunt 
in entirely satisfactory relationship. Subject thinks he is less 
talented than his brothers, one of whom is in training for the 
priesthood. 



290 SEX AND PERSONALITY 

H-3 attended only parochial schools and, like his brother, after 
graduating from high school entered a Catholic seminary and 
began preparation for the priesthood. After four years this plan 
of study was discontinued because of the opposition of his aunt, 
who was a Protestant. 

Subject is of large physical build, rather fleshy and voluptuous. 
He has large hips and breasts but a relatively small head and 
face. Has never taken part in athletics, in fact made every 
effort to avoid taking physical training when in school. 

Throughout his school life was treated as a sissy and preferred 
to play with girls "because they did not fight like boys." His 
favorite toys were dolls; favorite play, house: "I was just like a 
mother." Most liked school subject was chemistry; most 
disliked, French and Greek. 

After the discontinuance of his preparation for the priesthood 
the subject obtained, as the result of his interest in chemistry, a 
laboratory position in a small hospital. He became interested 
in caring for the sick and took a short course in practical 
nursing, and has been engaged as a male nurse for the last 
six years. 

At the time of the examination H-3 was a male prostitute in 
San Francisco. He had come there from Chicago in the com- 
pany of an aged invalid man. The old man died shortly after- 
ward. Subject claims to have been unable to procure work 
and that he prostituted himself only to make a living until he 
could get something else to do. 

At the age of 12 he was taught to masturbate by other boys. 
A short time before that his first homosexual experience occurred. 
While sleeping with his brother he recalls waking up and finding 
himself engaged in manustupration and attempting fellatio on 
him. The brother awoke about the same time and forced him to 
stop. It is interesting to note, however, that there was con- 
siderable rivalry between the subject's two brothers as to which 
should sleep with him. On a number of occasions one of them 
even paid for the privilege. 

First engaged in homosexual acts regularly while in the 
seminary, where he reports such practices to be very common. 



CASE STUDIES OF HOMOSEXUAL MALES 291 

He prefers fellatio, but will play the passive role in pederasty 
when it is demanded. 

He says he has never had a heterosexual experience, although 
women do have a slight amount of attraction for him. He 
enjoys his homosexual practices but feels that he would be much 
happier if normally sexed. Thinks that he might be able to 
change over if he could find a suitable girl to marry. Before 
coming to San Francisco he had been going regularly with a 
female nurse. 

H-3 began having nocturnal emissions at 17 and occasionally 
has them at the present time. The accompanying dreams are 
always of men and boys. 

Significant facts: favored child; both parents affectionate; loss 
of father when boy was 12; dose intimacy with mother; doll play 
and preference for girl playmates; segregation in seminary. 

H-4. "Lorraine" 

Exercise 1234567 Total 

M-F score -11 0+5-30-32-3 0-71 

"I" score -83 -27 +107 -173 +125 -47 -27 -125 

The M-F score is about 2.75 S.D. below the mean for high- 
school boys, and is close to the mean for high-school girls and 
women teachers. The "I" score is about 1.7 S.D. less invert 
than the mean of the P.M.H. group. 

H-4, age 19, is an attractive Mexican boy, small and plump. 
He was the third of five children, two of whom died in child- 
hood. Little is known of the father, as he died when the subject 
was 3 years old. Child was reared by mother and maternal 
grandfather. The mother, now 40, is very affectionate and 
sensitive. She and her son are extremely intimate and spend 
much of their time together. He says, "We are just like two 
sisters." He has assisted her much in the housework. 

There is one brother living and one sister. Although H-4 
greatly respects his brother, the relationship between them is 
much less dose than between the subject and his sister, whom he 
adores. 

His favorite childhood toys were dolls; favorite game, playing 
house with little girls. Here he always insisted on being the 



292 SEX AND PERSONALITY 

"mother." He was always afraid to take part in the rough 
games of boys, but he was much liked by adults because of his 
happy disposition. 

H-4 attended school as far as the seventh grade. He was 
considered a sissy by the other children and finally quit school 
because of ridicule. His favorite subjects were history, geogra- 
phy, and English; he disliked arithmetic. He liked to play 
volleyball but he "could not get along with the boys." 

Subject is a sociable person and spends most of his time in a 
group. He has no interest in reading. Most of his close friends 
are women because "they understand me better." He was born 
a Catholic, but has since become a Protestant. He says he has 
never had a love affair or even a date with a girl. He thinks 
that men are far better looking than girls. Ideal character, 
Ramon Novarro. 

H-4 has worked as a bellboy and as a bus boy for periods of 
one year each. He was also employed in a rubber factory for a 
year and liked the work. 

He obtained his first sex information "from friends" at about 
15 years of age. Shortly afterward he learned to masturbate 
and practiced the act about twice weekly. When 16 he met a 
"good-looking sailor" in Los Angeles who took him to the home of 
a male prostitute for the evening. It was there that his first 
homosexual experience occurred, when he was taught to play 
the passive role in pederasty. This practice has since been 
commercialized on numerous occasions. He has engaged in 
fellatio only a few times; he does not enjoy it and always feels 
ashamed of himself afterward. 

Women are sexually unattractive to him and he has never had 
a heterosexual experience in spite of many opportunities. Has 
slept with women without being the least aroused; feels that he 
would be ashamed if he should force himself to heterosexual 
intercourse. 

Subject says he has had only one nocturnal emission, which 
occurred about six months ago. The accompanying dream had 
a young man as the central character. 

Significant facts: early death of father, which probably 
strengthened the bond between mother and son; mother unduly 



CASE STUDIES OF HOMOSEXUAL MALES 293 

affectionate; fondness for sister; early preference for girls and 
girls' activities; small size and physical attractiveness: seduced 
at 16. 

H-5. "Dolores" 

Exercise 1234567 Total 

M-F score: -12 -1 + 9-62-1-3-70 

"I" score: +26 +23 +219 +146 +440 +135 +124 +1113 

The M-F score is about 2.75 S.D. below the mean for high- 
school boys and falls close to the mean for women teachers. 
Exercise 5 is especially feminine. The "I" score is highly 
invert, being one S.D. above the mean of the P.M.H. group. 

H-5, age 21, is an only child. The father is now 62 and the 
mother 52. The father was a marine engineer and was often 
away from home for long periods. He is decidedly masculine 
and quick tempered, though indulgent toward his son whom he 
has never punished. However, there was little companionship 
between them, as the father was so much older and so much away 
from home. The mother is described as well educated, extremely 
emotional, and the finest person the subject has ever known. 
Both parents wanted a girl when H-5 was born. 

As a child the subject preferred to be alone; he sometimes 
played with other children but did not enjoy them; was called a 
sissy by his schoolmates. 

Has always preferred girls' clothes and when still a youngster 
secretly dressed in them. Upon being discovered was made to 
wear girls' attire as a punishment and gloried in it. Played 
female roles in several school plays. 

H-5 completed high school and attended a western state 
university for one year. His favorite studies were English, 
French, and criminology. He disliked mathematics and 
chemistry. 

He dresses neatly and conservatively in dark colors, although 
he is fond of bright colors in ties, socks, shirts, and underwear. 
He smokes and drinks and has used morphine. He is a member 
of the Episcopal Church. 

From the age of 6 he greatly enjoyed seeing men's genitals 
and at 12 indulged in mutual fellatio with a fifty-year-old music 
teacher who was later sent to prison for his activities with other 



294 SEX AND PERSONALITY 

boys. The practice of fellatio continued for about four years, 
after which for two years he was inactive because of lack of 
opportunity and fear of being found out. 

At the age of 18 he met a group of " queens " and was fascinated 
by their life and adventures. They speedily took him into their 
group and introduced him to a sailor of whom he became enamored 
and with whom he claims he is still in love. 

He practices both pederasty and fellatio and prefers men 
about 30 who are "not too rough yet not sissified." Women 
have no attraction for him and he has never had a heterosexual 
experience. 

H-5 is attractive, intelligent, and interested in music. Studied 
piano for several years. 

Significant facts: an only child; father indulgent but much 
away from home; mother emotional and affectionate; parents 
wanted a girl; preference for girls and girls' activities; transves- 
tism; seduced at age of 12; scores highly "invert" as well as 
highly feminine. 

H-6. "Leslie" 

Exercises 1234567 Total 

M-F score: -3 -2-7 + 2-50-1-2-63 

"I 7 ? score: -42 +14 +154 +135 +275 +107 +148 +791 

The M-F score is about 2.6 S.D. below the mean of high-school 
boys and close to the mean for female nurses. Exercise 5 
(interests) is highly feminine. "I" score is about one-third 
S.D. more invert than the mean of P.M.H. group. 

H-6, age 19, is of very boyish appearance. He was the third of 
five children. The father, who died when subject was 4 years 
old, is described as moody, autocratic, cruel, and unstable. He 
was at various times a house-to-house salesman, a prizefighter, 
and a politician. The mother, now 43, is energetic and emo- 
tionally unstable. After the death of her husband she married 
her husband's brother, who developed a dislike for the step- 
children. H-6 could not endure being in the same room with 
his stepfather. Mother the favorite parent; the relationship 
between her and son very intimate; the mother gave in to the 
child and spoiled him. 



CASE STUDIES OF HOMOSEXUAL MALES 295 

H-6 seems to have had a normal childhood, except that he 
preferred to be with girls and to take part in their games instead 
of associating with other boys. He recalls that one of his 
favorite pastimes was dressing dolls and arranging them for 
funerals. 

Subject attended a Catholic school as far as the first year of 
high school. Quit because the work became monotonous and 
his parents had not enough money to send him further. Has 
read widely in all fields since leaving school and has a good fund 
of general information. He soon lost all faith in religion, and 
would often argue for hours on the philosophy of religion with a 
nun, a friend of his family. 

His feminine interests persisted after quitting school. He 
enjoyed doing housework and taking care of children. About 
this time, however, he developed a complex against girls and 
cared nothing more for their company, though he remained fond 
of his mother. 

Most of his friends were older men. He "despised" Boy 
Scouts and never cared for any kind of club activity. This 
antisocial attitude has continued; he prefers to live and sleep 
alone, and likes to take long walks alone. His interest in reading 
has not diminished. 

H-6 worked for short periods as a typist, waiter, office boy, and 
cashier in a tearoom. He says that he never enjoys any kind of 
work and will never take another job. He now earns his living 
solely by prostitution. 

He obtained his first sex information from books at about 
12 years of age. Masturbation, both alone and with other 
boys, occurred a short time later. Dreams of boys always 
accompany his nocturnal emissions. 

His first homosexual experience (mutual masturbation and 
pederasty) occurred at the age of 13, when he was seduced by an 
older man. No feeling of guilt followed this behavior. He has 
never practiced fellatio and regards the practice as disgusting. 

He has never had a heterosexual experience although a girl 
with whom he was sleeping when he was IS years old tried to lure 



296 SEX AND PERSONALITY 

him into coitus. He was disgusted and reports noting a dis- 
agreeable odor at the time. 

H-6 is timid in the presence of others and talks but little. He 
finds it almost impossible to urinate in the presence of others. 
He spends most of his time reading and only goes into a crowd 
in order to obtain the " business " by which he lives. He reports 
that the sexual act which follows is never pleasant and is often 
disgusting. Even so, he prefers it to any kind of work. 

Significant facts: early death of father, stepfather who disliked 
him; mother emotional and excessively affectionate ; spoiling; 
preference for girls; seduced at 13; "I" score decidedly invert. 

H-7. "Gypsy Q" 

Exercise 1234567 Total 

M-F score: -5 -1 - 19 -17 -8-2 -4-56 

"I" score: +3 -9 +161 - 1 +130 -36 +66 +314 

The M-F score is at the mean for gifted girls and about 2.5 S.D. 
below the mean of high-school boys. It is feminine for all the 
exercises. The "I" score is about two-thirds S.D. less invert 
than the average for the P.M.H. group. 

H-7, age 33, is a full-blooded Spaniard of average height and 
more than average weight. Not especially feminine in appear- 
ance. He is decidedly so in behavior, speech, and attitudes. Is 
extremely talkative and affected in conversation and manner- 
isms. Looks much younger than his age. Mentioned his "per- 
fect knees" as his most feminine feature. 

Subject was the second of seven children, but the first child (a 
girl) died when only a few weeks old. There are now living four 
brothers and one sister. The father was a merchant in a small 
Spanish city and died when the subject was about 14. He 
seems to have been of a stable disposition and kindly but stern. 
He was interested in the son but did not show much affec- 
tion. The mother, on the other hand, was extremely affectionate 
and soon became the favorite parent. She is well educated and 
is said to be descended from a prominent Spanish family. Both 
parents are Catholic; the mother is exceptionally devout and 
given to mysticism. 



CASE STUDIES OF HOMOSEXUAL MALES 297 

H-7 was the mother's favorite child. The father, however, fav- 
ored the younger sister, who was the only girl in the family. Since 
the firstborn baby, which died, was a girl, one wonders whether 
H-7 did not fill the deceased girPs role in the mother's affections. 

All through his childhood H-7 was treated as a girl both by his 
parents and by his playmates. He played mostly with girls 
because boys were too rough to be good playmates. His favorite 
toys were dolls and doll dresses; the favorite game was playing 
house. 

His mother taught him to read and he attended Catholic 
schools between the ages of 9 and 20, the last three years of which 
were spent in a seminary in preparation for the priesthood. He 
had entered upon this course at the urgent request of his mother, 
whom he could not bring himself to refuse. He claims to have 
enjoyed the study of philosophy, history, literature, biography, 
anthropology, theology, sociology, and psychology; he "hated" 
mathematics. 

At the age of 20, H-7 became attached to a ship officer who 
agreed to get him a job as mess boy on his ship if he would quit 
school and go with him. As he did not want to become a priest 
he gladly accepted the offer. The friendship for this officer 
gradually developed into love, which finally resulted in the prac- 
tice of fellatio and pederasty with him, H-7 always playing the 
passive role. The affair was finally broken up by the transfer 
of the officer to another ship and the young man found himself 
alone in Philadelphia. Being unable to speak English fluently 
and without special training for any occupation, he found it 
impossible to obtain suitable employment. He was too proud 
to work as a common laborer. The outcome was that he took 
advantage of his feminine characteristics and began playing the 
role of male prostitute, which proved so profitable that he con- 
tinued to follow it as a means of livelihood. Subject disdains 
any kind of work; insists that he will not sell his mental faculties 
or his physical energy for a mere $1.50 a day when it is easy to 
get $5.00 a day by prostitution. 

Obtained his first sex information from schoolmates and engaged 
in mutual masturbation with them at about 14. Masturbation 



298 SEX AND PERSONALITY 

is still regularly practiced in addition to his other sexual activity; 
it is carried out with narcissistic admiration of his own naked 
body. 

He has never had a heterosexual experience and women are 
utterly unattractive to him. During the last ten years he has 
been the passive member in several semipermanent unions with 
men. His ideal lover is large, muscular, and athletic, with 
congenial intellectual interests. Subject is very fond of male 
statuary. 

His first nocturnal emission occurred at 14 and others have 
occurred intermittently since then. They are accompanied by 
fantastic and exciting dreams but never include either men or 
women as a predominant feature. 

Considers himself a case of congenital homosexuality and is 
therefore unashamed; however, he looks upon normal men who 
patronize "queens" as degenerates. 

Significant facts: treated as a girl by excessively affectionate 
mother, apparently as surrogate for deceased older sister; 
segregation in a seminary for three years; lack of occupational 
training and low opinion of work. 

H-8. "lola" 

Exercise 1234567 Total 

M-F score: +7 -1-8+47 -80 0-1-50 

"I" score: -48 +18 +169 +125 +463 +89 +116 +932 

The M-F score is nearly 2.4 S.D. below the mean of high-school 
boys and dose to the average of Who's Who women. Exercise 5 
(interests) is especially feminine. "I" score is nearly three- 
fourths S.D. more invert than mean of P.M.H. group. 

H-8 is 20 years old, the youngest of nine children (four boys 
and five girls). Father died five years ago at age 60, mother a 
year ago at 59, both of heart disease. Parents were of Scotch 
descent. The father was a candy manufacturer, but was an 
invalid for several years before his death; the mother had a 
"good business head." There is no record of crime in the family. 

The parents wanted a girl at the time of subject's birth. As a 
child his favorite toys were dolls and dishes and his favorite game 
playing house. He played about equally with boys and girls, 



CASE STUDIES OF HOMOSEXUAL MALES 299 

but was always considered by the boys as "different." His 
favorite school subjects were English, spelling, and Latin; he 
disliked mathematics. 

H-8 was a constant companion to his father during the latter's 
illness and so led a rather quiet and secluded life, spending most 
of his time in reading. The father cared more for him than did 
the mother. 

While still at home, the subject liked to dress up in girls' 
clothing but was not allowed to do so. After leaving home at the 
age of 17 he dressed as a woman for three months before being 
detected. He still attends masquerades in women's clothing at 
every opportunity. Recently had much newspaper publicity 
from masquerading as a Russian princess. Usually, however, 
he dresses as a man in order to evade the police. 

He once worked for five weeks as a waitress (sic.) and later for 
two months as a clerk in a mine office. 

He is of a Mormon family and is still a nominal member of 
this church. His favorite amusements are concerts, shows, and 
books. He drinks to some extent but does not smoke. 

H-8 first masturbated at the age of 15, having been taught by 
an older man. The act was practiced about five times a month 
until other means of gratification were found. His first hetero- 
sexual experience occurred at the age of 16 when he visited a 
brothel with some school companions. He first committed 
pederasty at the age of 17 with a boy of the same age, and reports 
that he did so because he wanted to. This practice has been 
commercialized and has been practically his sole means of support 
for the last three years. He first practiced fellatio about six 
months ago while partially intoxicated. 

Significant facts: parents wanted a girl; early preference for 
girls' activities; companionable with father (unusual in these 
subjects), perhaps accounted for by the parent's invalidism; 
transvestism. 

H-9. "Rose Marie" 

Exercise 1234567 Total 

M-F score: -IS -8 +6 -26 +3 +3 - 37 

"I" score: +964 



300 SEX AND PERSONALITY 

M-F score is about 2.2 S.D. below the mean for high-school 
boys, and far more feminine than the mean of women athletes. 
"I" score total is about three-fourths S.D. more invert than the 
mean of P.M.H. group. 

H-9, age 31, is the only living child in a family of ten children. 
All the others were born dead or died in early infancy. The par- 
ents were remote cousins. The family is reported to be in good 
circumstances and socially prominent in a middle-western city. 

The father, now 68, has been a physician for 25 years. He is 
said to be very stable, stern, and intelligent, and has a good 
medical practice. He was autocratic toward the boy but never 
cruel. The mother, now 62, seems to be extremely emotional 
and unstable. She was the favorite parent, because "she 
would always stick up for me and give me anything I wanted." 
On one occasion she locked the child in a doset as punishment, 
but at the end of 15 minutes let him out, gave him $2.00, and in 
tears begged his forgiveness for her cruelty. The two were 
together constantly, the mother was demonstrative in her affec- 
tions, and the boy worshiped her. 

Both parents are of French descent, Anglo-Catholic in religion, 
and well educated. They tried to rear their only child properly 
but instead spoiled Him and perhaps laid the basis for his unhappy 
and unsuccessful life. They gave him all kinds of toys, books, 
and games, and tried to make of him a model child. Educated 
in a private school he was not allowed to play with the other boys 
of the neighborhood and was always accompanied by a negro 
servant when he went out of the yard. Was permitted to have 
girl playmates and he enjoyed playing girls' games. His favorite 
toys were dolls, although he had expensive mechanical toys for 
which he did not care. His favorite school subject was history; 
grammar was disliked. 

H-9 is now rather social, but his favorite associates are "queer 
people and the intelligentsia." He is a member of three fraternal 
organizations, but he joined them only because his father insisted 
that it was the thing to do. He was a member of the Episcopal 
Church but reports that he cared only for the "show"; he acted 
as an altar boy for a time and liked the part. 



CASE STUDIES OF HOMOSEXUAL MALES 301 

Subject obtained his first sex information at the age of 14 from 
an uncle. Soon he was practicing mutual masturbation and 
fellatio with a boy friend and attempted sodomy with him. At 
first these practices were engaged in solely for sexual gratification, 
but they became quite pleasant and a feeling of love for the other 
boy developed. 

His father tried in vain to get him to work in a drugstore and 
his mother attempted to persuade him to become a priest. 
Neither course appealed to him and he ran away to Chicago. 
There, learning that he could make an easy living by com- 
mercializing his sexual practices he entered a "peg house." 1 
He stayed there for about two years before deciding to return 
home to enter college as his parents wanted him to do. 

After a French custom, the parents had betrothed him as a 
child to one of his second cousins. While at home on this visit 
his father insisted that he get married. The girl he married was 
reserved and cool, almost frigid. Subject slipped away after 
the ceremony and did not return for three days. He then 
decided to settle down, become as masculine as possible, and be 
a good husband. His first heterosexual experience occurred 
shortly afterward with his wife, but he found it disgusting. 
Three months later, on learning that his wife had become preg- 
nant, he ran away from her and went back to the peg house in 
Chicago. He never returned to his wife but has learned indi- 
rectly that he has a son of 11 years. The wife has never 
remarried. 

After a short stay in Chicago H-9 began to roam around all 
over the country, always making a living by prostitution. Says 
he has made a large amount of money in this practice and that 
he was unusually successful on the "Barbary Coast" of San 
Francisco during the pre-Prohibition period. 

He fell in love with a young German bootlegger in Detroit 
and while posing as a woman was legally married to him. They 
were happy together until the "husband" was killed in a gun 
fight with the police. The subject suffered greatly from shock 

1 House of prostitution for boys. 



302 SEX AND PERSONALITY 

and grief, decided to abandon prostitution, and took a course in 
nurse's training. Since completing this course he has been 
employed almost solely as a male nurse or hospital orderly. 

His homosexual practices have continued, however, and are 
still resorted to as a means of livelihood when he has no work. 
When last heard of he was still "on the street." 

Significant facts: accentuated case of "only" child; parents 
overindulgent, especially the emotionally unstable mother; cut 
off from contact with boys; lack of occupational training; "I" 
score more invert than average of P.M.H. group. 

H-10. "Aimee" 

Exercise 12 34567 Total 

M-F score: -16 +4+4-8 -10 - 2 - 28 

"I" score: -62 +1 +124 +23 +184 +61 +41 +373 

The M-F score is two S.D. below the mean of high-school boys 
and more feminine than the mean of women athletes. The "I" 
score is more than .5 S.D. less invert than the average of the 
P.M.H. group. 

H-10, age 23, is the fourth of a family of six children, four boys 
and two girls. He is of medium build, somewhat plump, and 
fairly attractive. He has a soft effeminate voice and tends to be 
rather quiet. 

The father, age 65, is a veterinarian. He seems to possess no 
outstanding characteristics. There was considerable intimacy 
between father and son, but also fear on the son's part. The 
mother is now 53 and a housewife; she worked for a time as a 
dressmaker. She too is reported as not having any marked 
peculiarities. The mother was the favorite parent because 
subject "had been with her more." When pressed for the reason 
for their constant companionship the subject simply replied that 
he and his mother "seemed to understand each other." He 
reports that as a child he worshiped her and helped her a great 
deal in her work. 

The mother wanted a girl at the time of his birth and even 
prepared for a daughter. The subject was also the favorite of 
the grandmother who lived with the family and made over him 



CASE STUDIES OF HOMOSEXUAL MALES 303 

a great deal. An uncle is thought by the subject to be a 
homosexual. 

As a child the subject was considered a sissy by his playmates 
and treated by them more like a girl than a boy. He preferred 
to play alone rather than in a group. His favorite toys were 
trains and dolls; his favorite games, checkers, puzzles, and house. 
He was most fond of English and mathematics and found manual 
training very difficult. He graduated from high school, but had 
no desire for further education. 

The only job held by the subject for any length of time was that 
of clerk in a grocery store. This position he held for two years 
after leaving high school. Since that time he has earned a living 
by odd jobs and prostitution. 

He is a Catholic but has never been very devout. Contrary 
to his childhood habits, he is now socially inclined and spends a 
large amount of time in the company of his friends. 

His early sex information was acquired from a middle-aged 
owner of a pool hall, and it was with him that his first sex experi- 
ence (mutual masturbation) occurred. Men have always been 
sexually more attractive to him than women, but he has also 
been attracted by a few girls. His first strictly homosexual 
experience occurred at age IS or 16, when he and a schoolmate 
with whom he was in love practiced mutual fellatio. He played 
the passive role in sodomy a short time later with the same boy. 
Subsequently he engaged regularly in both practices on a 
commercial basis. 

At 18 he was living in a private family and became somewhat 
interested in one of the girls. It was with her that his first 
heterosexual experience occurred. The practice was con- 
tinued regularly during the remainder of his stay with the family 
(about a month). The subject also has occasional heterosexual 
experiences at the present time but claims that they are not 
nearly so satisfying as homosexual experiences. 

He had his first nocturnal emission at the age of 14 and usually 
dreams of men rather than women on such occasions. 

He feels that homosexuality is the result of prenatal influences 
and therefore not subject to change. 



304 SEX AND PERSONALITY 

Significant facts: mother wanted a girl; mother the favorite 
parent, but no antagonism to father; early preference for girls; is 
to some extent bisexual. 

H-ll. "Myrna H 

Exercise 1234567 Total 

M-F score: +4 -8-5-16+1 +1-23 

"I" score: +2 +5 +133 -64 +109 -36 -73 +76 

The M-F score is nearly 2 S.D. more feminine than the mean 
of high-school boys, and appreciably more feminine than the 
mean of women athletes. The "I" score is relatively low. 

H-ll is a diminutive youth of 23 whose whole physical make-up 
is decidedly feminine. Is oneof the best examples of the "queens" 
in the group. Both parents were English. His father left home 
before subject's birth and nothing is known concerning him. 

H-ll was a premature child, and continued to be sickly until 
6 years of age, spending a great deal of this time in hospitals. 

He lived alone with his mother, an emotional woman, until he 
was 18. He was brought up to do housework and all other 
things commonly done by girls. His favorite toys were dolls, his 
games were those of girls, and he strongly disliked boyish sports. 
He liked writing, reading, and history, but disliked arithmetic. 
He played all the time with girls although there was ample 
opportunity for association with boys. He has a picture taken 
at the age of four in which he appears in a dress and long curls. 

At various times he has collected stamps, signatures, and 
girls' pictures. He is very fond of parties and dances, especially 
when he has an opportunity to play the feminine role. 

He has always sought work which would enable him to be with 
girls. He enjoys their companionship, but is never sexually 
attracted toward them. He had one female sweetheart who 
did not like the way he acted toward her "not like other 
fellows." He has never had a heterosexual experience nor any 
desire for one. 

When arrested (he is now in prison) he had been working for 
five months as an instructor in a girls' gymnasium. He has 
marked religious inclinations, having been a member of a great 
many denominations at one time or another. 



CASE STUDIES OF HOMOSEXUAL MALES 305 

He first masturbated at 14 but had before then felt very 
pleasurable sensations on looking at men doing manual labor. 
His first homosexual experience occurred at 18 with a handsome, 
athletic fellow, an ex-college student. He denies ever having 
practiced fellatio, which he says is nauseating to him, but admits 
an extreme liking for sodomy provided he is allowed to play the 
passive role. He is always the one pursued. 

He has masqueraded as a woman whenever possible and says 
that he would give anything in the world to be able to dream of 
motherhood. H-ll is passionately fond of babies and con- 
sistently calls himself "she." Partly because of the fact that he 
weighs 100 pounds and is only 5 ft. 2 in. tall, he is very popular 
as a "queen" among the rest of the convicts. Says he has 
had 75 love affairs in the 18 months that he has been in prison. 

A deep (though soft) voice is his outstanding masculine 
characteristic. 

Significant facts: mother was deserted by husband before 
child's birth; sickly childhood; treated by mother as a girl; 
transvestism; feminine physique; "I" score is relatively low. 

H-12. "MissX" 

Exercise 1234567 Total 

M-F score: -18 -1 -7 +17 +12 -1 -2 +00 

The M-F score is nearly 1.5 S.D. below the mean of high-school 
boys. Among high-school and college men about one in ten is as 
feminine. The score coincides almost exactly with the mean for 
artists. 

H-12, age 24, is slightly below average in height and weight, 
attractive, and courteous. He is an only child. His father was 
a stock-and-bond salesman and died when the subject was only 
3 years old. His work kept him from home most of the time, and 
subject has practically no memory of him. The mother died 
two years ago. She is said to have been gentle, industrious, and 
much less emotional than the average woman. She worked 
as a housekeeper after her husband's death. 

The death of the father resulted in a very dose relationship 
between mother and son, although there was but little display of 



306 SEX AND PERSONALITY 

affection on the part of either. The relationship was more like 
that between an older and younger sister, and involved constant 
companionship. The boy helped his mother with the house- 
work and soon became quite efficient. 

H-12 developed into a sissy but an unusual one in that he 
disliked girls. He associated with whatever boys would tolerate 
him and tried to play their games. He did not care for dolls; was 
very fond of electrical and mechanical toys. His favorite games 
were tiddlywinks and ping-pong. Cared but little for baseball. 
Liked to be with older people. 

Subject quit school at the end of the seventh grade because of 
his mother's illness. His favorite studies were history, geography, 
and reading; arithmetic was thoroughly disliked. Since leaving 
school, he has spent a great deal of time in libraries. 

When asked how he differed from other children of the same 
age and sex, subject replied, "I daydreamed a lot, hated vulgar 
language, liked to argue, and did not believe in religion. My 
playmates considered me slightly freaky." 

When still a boy he was treated by adults more like a grown 
person than a child. He says that he knew everyone in town;- 
preachers, businessmen, and loafers were all his friends. They 
were kind to him, kept him in spending money, took him to shows 
and dub meetings, and made him feel that he was one of them. 
He attributes this generosity to the fact that his father was a 
thirty-third-degree Mason. 

At the present time H-12 is quite social and enjoys being 
among sympathetic friends. He belongs to the Y.M.C.A. but 
to no lodges or dubs. 

He worked as a hospital orderly for four years, but quit 
because he received no advancement. He has also been employed 
for varying lengths of time as a bus boy, factory worker, and 
houseboy, but he does not like to work; thinks he might enjoy 
being a male nurse, but since he makes plenty of money as a 
prostitute there is no need to change. 

Women are sexually unattractive to him. He has never had 
heterosexual experience in spite of the fact that he once worked 
several weeks for a female prostitute. 



CASE STUDIES OF HOMOSEXUAL MALES 307 

Subject obtained his first sex information at the age of 14 from 
other boys and began to masturbate at that time. Still mas- 
turbates at irregular intervals. 

Although H-12 claims that he remembers awaking in the act 
of kissing a friend's penis at the age of six, his first adult homo- 
sexual experience occurred at 16, when he played the passive role 
in fellatio with a middle-aged man who persuaded him to do so. 
Subject allowed the same man to practice pederasty on him a 
short time later. Fellatio has remained his favorite practice; 
pederasty is engaged in only at the request of his patrons. 

Nocturnal emissions occur but rarely and they are accom- 
panied by dreams of boys. 

Subject says he would rather be a homosexual than change 
over even if he could. He claims to have monthly backaches 
and headaches which he fancies are similar to menstrual pains. 

Following is a short note written by the subject: 

I remember when I was at the age of 13, some friends of mine had as 
guest a man of 32 years who came for a visit to their house. One day 
while he was bathing I came into his presence with a towel and would have 
actually refused to leave if I had been requested to, as I was all a-tremble 
with passion. I stayed on and offered my assistance. It was laughingly 
accepted. I helped in rubbing his back and I could hardly keep myself 
from kissing his buttocks after I was through. Although knowing he had 
rubbed himself dry in front, I had to do it all over myself for him; I 
lingered so long over him (and must have actually begun to caress his 
body with my hands) that he began to have an erection and tried to put 
it in my face. I refused to accept it, although I wanted to. 

Significant facts: absence and early death of father; close 
association with gentle and kindly mother; a sissy but disliked 
girls; liked mechanical toys; disliked vulgarity; received favored 
treatment from adult males; seduced at 16, though homosexual 
tendencies were evident before this age. 

H-13. "Florrie" 

Exercise 1234567 Total 

M-F score: -13 +1 4-16+40-30 -4 -3 + 7 

"I" score: -49 +8 +122 -40 +258 +5 +57 +345 

The M-F score is about one and a quarter S.D. more feminine 
tlian the mean of high-school or college men, is within the 10 



308 SEX AND PERSONALITY 

per cent most feminine of men in general, and is dose to the 
mean for the male theological group. The "I" score is about 
three-fourths S.D. less invert than the mean of the P.M.H. 
group. 

H-13 is a fairly attractive red-headed boy of 23, with effeminate 
voice and mannerisms. He was the first child of Irish-American 
parents who are now engaged in the restaurant business. His 
mother, age 46, is of an emotional type as contrasted with the 
father, age 52, who is cold and unresponsive. Because of incom- 
patibility the parents separated and were divorced. Both 
married again but the new partners in each case died and the 
parents are remarrying. 

A sister, age 20, is a waitress in her father's restaurant. She 
is very sympathetic with H-13 in his failings. A brother, killed 
by an auto at the age of 11, was much loved by the subject. 

Because of her husband's dislike for children the mother did 
not tell him of her pregnancy for nearly four months. She 
wanted a girl, while the father hoped that the child would be 
born dead. 

All through childhood the mother was most affectionate with" 
her son. He loved her intensely and feared his father. She 
made a special point of teaching him to be neat with his clothes, 
a trait that is still outstanding. 

As far back as H-13 can remember he liked to play with girls, 
wear feminine clothes, and help with housework. He boasts of 
being an excellent cook and housekeeper. On every possible 
occasion he would steal his mother's or sister's clothes and dress 
up as a girl. He says that he did this because he knew that he 
looked better in girls' clothes. He recalls that he used to dress 
as a girl and turn somersaults at five cents a "turn" for the 
laboring men who worked near by. On one occasion he stole a 
brilliant red dress belonging to his aunt and wore it to the funeral 
of a neighbor whom he disliked. Apparently his impersonations 
were successful. He frequently took the feminine role in plays 
and enjoyed acting. 

At school he received good marks, doing especially well in 
English literature, composition, art, and music. He always 



CASE STUDIES OF HOMOSEXUAL MALES 309 

contrived to be the teachers' pet and on a number of occasions 
was taken home overnight with them to be paraded as a model 
pupil. His grammar and vocabulary are above average. He 
disliked mathematics and cared little for spelling. 

H-13 has a good voice and says he was once known as the best 

soprano in (a large midwest city). On several occasions he 

has considered joining a vaudeville troupe and posing as a female 
singer. 

Subject has spent most of his time working in a restaurant but 
was employed as a hospital orderly for several months. He is 
sympathetic and enjoys caring for others. 

His favorite pastimes and hobbies are reading, listening to good 
music, and seeing good shows. He is also fond of travel, a fact 
which he attributes to his restless disposition. 

Nominally he is a Catholic but his wide reading has rid him 
of all orthodoxy. 

H-13 early noticed that he experienced pleasure at the sight of 
strong muscular men; would "sit for hours and watch a brick- 
layer or a blacksmith at work." When at IS he joined the navy, 
he found that it was practically impossible to ward off the atten- 
tions of his fellow sailors, as they were attracted by his feminine 
qualities and pursued him constantly. He would move into a 
new place in order to be alone and immediately two or three men 
would move in with him. At that time he wanted only to love 
and kiss his companions, and had no desire for sexual congress. 
Before long, however, he was forced into playing the passive 
role in sodomy and was also taught fellatio. For a time these 
practices were followed by a feeling of shame, but after a time 
sodomy became a pleasant experience and developed into a 
habit which has remained his chief type of perversion. He does 
not practice fellatio, but on request has sometimes played the 
active role in sodomy. 

When he was 17 he became infatuated with a young girl whom 
he met at a navy dance and had with her his first heterosexual 
experience. However, he rarely finds a woman for whom he 
has any desire, and then the satisfaction is only temporary. He 
says that he has sometimes to imagine his partner a man before 



310 SEX AND PERSONALITY 

he can succeed in the sexual act. He has never been married 
but has lived with a number of men for two or three months at 
a time. 

H-13 has a very changeable disposition, suggesting a typically 
unstable personality; he makes friends easily. He has an exces- 
sive number of nocturnal emissions, which he considers the result 
of his extreme emotionality. With these he dreams of both men 
and women, himself playing the part now of one, now of the other. 
He has many dreams of motherhood, of labor pains, and of 
menstruation, with himself as the chief character. 

Being above average in intelligence, H-13 is much interested in 
his case and is not at all reluctant to talk about himself. He feels 
that homosexuals as a group are misunderstood and mistreated. 
Regards his feminine tendencies as the product of inborn charac- 
teristics and early training by his mother. (About the only 
subject in the P.M.H. group to suggest any influence of environ- 
mental factors.) 

Significant facts: cold father, sympathetic mother; mother 
wanted a girl; neatness of person overemphasized; early and 
persistent tendency to transvestism; musical; seduced at IS; to- 
some extent bisexual. 

H-14. "Jeanette" 

Exercise 1234567 Total 

M-F score: + 2 +1 -3 +35 -20 -8 +1 + 8 

"I" score: +18 -24 +154 -37 +237 +30 -22 +356 

Both M-F score and "I" score almost identical with those of 
H-13. 

H-14, age 37, is an interesting case in that he has practiced 
homosexual acts for more than 20 years. He was the third child 
of a Dutch contractor and a Scotch-Irish mother. His father is 
described as very masculine and quick tempered, his mother as 
above average mentally and not especially affectionate. Both 
the other children were girls. One of them died at an early age, 
and H-14 grew up in the company of his remaining (older) 
sister, who is now a spiritualistic healer. He says that his mother 
always treated him much better than his father, who was brutal 
and "did not care what became of us." 



CASE STUDIES OF HOMOSEXUAL MALES 311 

The subject's childhood seems to have been normal except that 
his favorite toys and games were those of girls (dolls and playing 
house). He also took an interest in boys' games, however, and 
recalls playing ball and flying kites with much joy. In school 
he played chiefly with boys and claims to have been considered 
a roughneck among them. On the other hand, he liked to wear 
girls' clothes and has attended many masquerade balls dressed 
as a woman. 

H-14 is below average in height, but otherwise of average build. 
He has a markedly soft, effeminate voice almost gurgling at 
times. He dresses conservatively, smokes and drinks, and 
attends shows, concerts, and baseball games. He is a member 
of the Spiritualist Church. 

Subject has spent 16 years as a cook and waiter, but once 
worked in a machine shop for two years. At the present time 
is working as a night clerk in a small-town hotel. 

He says that he first masturbated at the age of 10 and con- 
tinued the practice until about 15 or 16, when his first hetero- 
sexual experience occurred. A short time later an older school 
companion induced him to play the passive role in pederasty, a 
practice which he has continued throughout his life and which he 
has often used as a means of livelihood. He says of his first 
experience, "When we were finished I realized that I had always 
wanted it, and I have continued to want it." He also practiced 
fellatio while a youth, but does not care to engage in it now. 
Prefers a lover who is athletic. Occasionally goes to a prosti- 
tute, but only when he is with male companions and accompanies 
them as a matter of sociability. 

Significant facts: brutal father; affectionate mother and older 
sister; liked both girls' and boys' games; transvestism; seduced 
at 15 or 16; to some extent bisexual. 

H-15. "Bella" 

Exercise 1234567 Total 

M-F score: -11 - 2 + 6 + 47 - 18 - 8 -3+11 

"I" score: +12 +28 +219 +194 +399 -14 +93 +931 

In M-F score rates dose to H-13 and H-H, but has an "I" 
score considerably more invert than the mean of P.M.H. group. 



312 SEX AND PERSONALITY 

H-15 is 33 years of age and the youngest of a family of eight 
children: four boys and four girls. He was born and educated 
in Germany, where he finished the Gymnasium at the age of 15 
years, 6 months, about 1J^ years younger than the average. 
From a very early age he showed feminine characteristics, such 
as preferring girls' games and toys and playing with girls almost 
exclusively. Upon receiving a train or other mechanical toy for 
Christmas he would immediately trade it to his tomboy sister 
for her doll. 

He was taught the practice of fellatio at the age of 12 by a 
German army officer of high rank. After leaving school he 
became an apprentice seaman on a German sailing vessel. He 
found this life distasteful and decided to remain in the United 
States. As he found himself attractive to men and popular with 
them, he drifted into the life of a male prostitute. Has held jobs 
as ship waiter, cook, and bellhop, but his chief source of income 
has been prostitution; is still an active "hustler." 

H-1S has a well-formed body, a pretty face, and a delightfully 
pleasant voice. He has a quiet disposition, speaks excellent 
English, and is an engaging conversationalist. Distinctly above 
average in intelligence. 

One brother is also a homosexual. 

Significant facts: youngest child; preference for girls' activities; 
seduction at 12; attractiveness to men; superior intelligence and 
culture; "I" score above average for the group. 

H-16. "Ruby" 

Exercise 1234567 Total 

M-F score: -4 -1 +9 -30+40 +3 -3 +14 

"I" score: +72 +17 +28 -157 -145 -27 +117 -95 

The M-F score is more than one S.D. below the mean of high- 
school or college men and close to the mean for male students of 
music. The minus M-F score on Exercise 4 and the plus score 
on Exercise 5 are both exceptional for this group. The "I" 
score is 1.6 S.D. less invert than the mean of the P.M.H. group. 

H-16, age 20, is the only son of Welsh parents. His mother 
was killed in an accident when he was 1 year old; his father was 
killed in 1917 while a soldier in Shanghai. The father is reported 



CASE STUDIES OF HOMOSEXUAL MALES 313 

to have been a heavy drinker, quite mean when intoxicated, and 
sexually promiscuous. An uncle on the father's side is serving a 
prison sentence for robbery. 

The subject was reared by an aunt and uncle, who were kind 
and affectionate toward their charge. He seems to have been a 
normal boy, playing the usual boys' games and taking delight in 
masculine toys such as trains and horses. At school he always 
played with the boys and took part in all branches of athletics. 
He completed a high-school course with printing, typing, and 
mathematics as his favorite subjects, but with a distinct dis- 
like for chemistry and biology. 

As might be expected in the light of his other interests, H-16 
preferred boys' clothes, and complained that "they made me 
wear short pants until I was nearly grown." It is interesting to 
note, however, that on two occasions he attended masquerade 
balls in female attire. 

He has never followed any occupation very long; when arrested 
(this case was tested in the San Francisco County Jail) was 
charged with vagrancy. Has worked for brief periods (3 to 17 
months) as a seaman (steward's helper), printer, waiter, laborer, 
and bootlegger. 

At the present time, his interests are typically masculine. He 
smokes a great deal and drinks. Is a member of the Catholic 
Church but attends rarely. 

A chum taught him to masturbate at the age of 10, and his 
first heterosexual experience occurred at 13 with a girl friend. 
Subject claims that since that time he has practiced coitus with 
women at least three times a week. He was married in Seattle 
in 1927 but was divorced six months later. 

He first engaged in pederasty with a German whom he had 
"quite a bit of love for." This was after he was divorced from 
his wife, at a time when he says that he was pretty well disgusted 
with women. Since this first occurrence, he has practiced peder- 
asty (passive role) regularly as a male prostitute. 

The subject is somewhat below average in size and very boyish 
in appearance. As his earlier interests seem to have been entirely 
heterosexual, it would appear that he is taking advantage of his 



314 SEX AND PERSONALITY 

feminine physique and simulating homosexual behavior in order 
to earn an easy living. The M-F score of +14, while low, is well 
within the normal range. 

Significant facts: normal masculinity as child; early and con- 
tinued heterosexual! ty; marriage and divorce leading to trial of 
homosexual practices; commercial motivation to continuance; 
M-F scores on Exercises 4 and 5 contrary to the rule for passive 
homosexuals; "I" score indicates no inversion. 

H-17. "Vardis" 

Exercise 1234567 Total 

M-F score: + 2 - 1 0+38-26 +8 - 3 + 18 

"I" score: -33 +28 +120 +18 +244 -6 +172 +543 

M-F score is one S.D. below the mean of high-school boys, close 
to the mean for fathers of gifted children, and definitely more 
masculine than the mean of artists and theological students. 
However, the "I" score is only a little less invert than the mean 
of the P.M.H. group, and it is this score, rather than the M-F 
score, that is in line with the subject's "queen" characteristics. 

H-17, age 22, is the only child of an unknown father, said to be 
a Hindu, and a negro-Indian mother. He is exceptionally neat 
in his appearance and has attractive eyes. 

Subject was reared chiefly by his grandmother, a very strict 
Christian, although his mother visited him from time to time and 
helped in directing his training. Mother, emotional and tem- 
peramental, very much wanted a girl at the time of his birth. 
He was never allowed to play with the "rough boys" of the 
neighborhood, and so took to housework, sewing, and playing 
with dolls. Played almost entirely with girls because "the 
boys were too uncouth and diabolical." He has a good soprano 
voice and is an excellent singer. 

He has worked as a chef, but his chief occupation has been 
that of a female impersonator on the stage; has followed this at 
intervals since the age of 13, usually playing the part of a South 
Sea Islander or a negro girl. He has been successful in this type 
of work and enjoys it. He has performed on the stage both 
locally and abroad, and has also played in the movies. He 
designs his own costumes, and while in prison (where he was 



CASE STUDIES OF HOMOSEXUAL MALES 315 

tested) spends most of his spare time in sewing and doing fancy- 
work. He has used tobacco and alcohol only occasionally, but 
was addicted to the drug habit for two years and was arrested 
in connection with his use of drugs. 

H-17 first masturbated at 14, although he remembers prac- 
ticing manustupration from the time he was 8. His homosexual 
experience also began about that time, sodomy being his first- 
learned perversion. He always plays the passive part in this 
relationship but stressed the point that he practices it only with 
those he loves. He detests fellatio and has no use for homo- 
sexuals who engage in the practice. 

He is fond of women as companions and is popular with them, 
but they have no sexual attraction for him. He says that he has 
never had a heterosexual experience. 

Like most of the others of this group, the subject believes that 
homosexual tendencies are inborn and for that reason not to be 
ashamed of. Says he has had no homosexual relations while 
in prison because he cannot find suitable partners; considers 
himself superior to others of his kind among the convicts. 
Although he reports caring most for quiet and refined men, he 
lived for several years with a prize fighter. 

Significant facts: reared by grandmother and mother; mother 
wanted a girl; was not allowed to play with boys; transvestism; 
early seduction; actor and singer; possesses exotic attraction to 
homosexual males. 

H-18. M. H. 

Exercise 1234567 Total 

M-F score: -21 +1+4+31 +8 -2 -1 +20 

"I" score: +8 -20 +138 +27 +60 -7 -7 +199 

The M-F score is practically the same as that of H-17, though 
more feminine in Exercise 1 and more masculine in Exercise 5. 
The "I" score is more than one S.D. less invert than the mean 
of the P.M.H. group. 

H-18 is 44 years old. Although in prison on the charge of 
rape, he is known to be practicing homosexual acts. 

Subject was the younger of two children, but his sister died at 
the age of 5 ; he was then 3. His parents were English-American 



316 SEX AND PERSONALITY 

and seem to have had no unusual personality traits. His mother 
died when he was 7, and from that time on, although a white 
child, he was reared by an old negro couple. 

He recalls wearing dresses as a small child (does not remember 
exact age) and that he did not like to play with boys because of 
their roughness. He was taught to be neat and to abhor all 
that was harsh and ugly. He had practically no toys and played 
very few games, as he preferred to spend his time alone with 
books and music. 

He began taking piano lessons at the age of 6, was playing a 
church organ at 11, and began giving piano lessons at about 13. 
He later established a studio and has spent his entire life in 
teaching piano, violin, voice, and dancing. He has also written 
a few mediocre musical compositions. 

He admits having had intercourse with one of his pupils, a girl 
of 14, but insists that they were planning to be married. How- 
ever, several girls testified at his trial that he had attempted the 
same offense with them. 

He first masturbated at about 16, and has visited prostitutes a 
few times. He strongly denies homosexual activities although 
he has been observed in the prison to practice sodomy, fellatio, 
and manustupration within the course of a few hours. 

H-18 has been in prison for about seven years and at the present 
time is extremely nervous perhaps as a result of his attempt to 
conceal his abnormal conduct. He tries to give the appearance 
of a goody-goody by feigning to be shocked with the vulgar talk 
of the other convicts. He insisted that he knew nothing of the 
homosexual activities of the other prisoners. He talks much 
of his former high social standing and of his many influential and 
socially prominent friends. 

Nominally he is a Protestant with definite leanings toward 
Christian Science. Although somewhat matronly in appear- 
ance, he has only a few specific feminine characteristics. His 
voice is soft and rather deep (he sings baritone) . His speech and 
carriage are noticeably effeminate, but not so pronouncedly as 
are those of many true homosexuals. 



CASE STUDIES OF HOMOSEXUAL MALES 317 

The following are excerpts from a statement about M. H. 
written by a fellow convict: 

M. H. likes to display himself naked and sleeps naked when possible. 
Has been reprimanded several times for not keeping dressed. Displays 
particular fondness for wearing surgical gowns that open in the back, 
wearing them without other clothing when possible. . . . 

When under the influence of sexual desire is excitable, abnormally 
nervous, shows difficulty in breathing and is apt to be particularly violent 
toward anyone who distracts his attention from his prospective lover. 
With "prospect" is very loving; caressing, fondling, and kissing with 
little caution about being seen. Very passionate, almost insatiable in 
his erotic desires. Attracted mostly by young fellows of any race or color; 
Mexicans, negroes, or whites, all are acceptable. . . . 

Significant facts: parents not unusual; an only child after age 
3; orphan; little contact with either girls or boys; personal 
neatness overemphasized; precocious in music; bisexual; violent 
passion and lack of inhibition; "I" score much less invert than 
the mean of his group. 

SUMMARY AND CONCLUSIONS 

The regularity with which certain items are repeated in the 
case histories of this group is very impressive. There are 
individual differences, but the composite picture is probably less 
blurred than it would be for almost any other male group one 
could find. 

Thirteen of the 18 were either only children (5), youngest (4), 
oldest (3), or sickly (1). 

Seven of the fathers either died or were away from home when 
the subject was very young. Eight were described as cold, 
stern, autocratic, or fear-inspiring; only three as kind or sym- 
pathetic. No father was the preferred parent. 

Six mothers were described as excessively affectionate and none 
as antagonistic or unkind; eight as emotional or unstable, one as 
markedly stable. Only one mother died when the subject was 
young and one other did not live constantly with the child. In 
eight cases the relationship between subject and mother is 



3 1 8 SEX AND PERSONALITY 

described as very close or intimate, in no case as distant or 
casual. 

Four of the mothers are said to have wanted a girl at the time 
of subject's birth; five are specifically said to have treated the 
child like a girl; that the number who did so was probably 
greater is indicated by the fact that six of the subjects helped a 
great deal with the housework. Two subjects mention special 
training in personal neatness. 

In four cases a grandmother enters into the picture; in three 
cases an older sister. 

Half the subjects showed marked preference for the com- 
panionship of girls, only one for the companionship of boys. 
Two had no opportunity to associate with boys and three 
preferred to be alone. Six report that they were regarded by 
boys as sissies. Normal male companionship has therefore been 
pretty well out of the picture. 

Only two of the subjects (one of these not a true homosexual) 
preferred boys' playthings; whereas, 11 were fond of dolls and 
"house" was the favorite play of five. 

Six have a history of transvestism, four or five of these being 
rather extreme cases. 

The average school grade completed was the third year of high 
school, which is significantly above the mean for males of their 
generation in any part of the United States. It is probable, 
therefore, that the group is also above the average of the general 
population in intelligence. 

As the most disliked study, mathematics (or arithmetic) was 
named by eight subjects, one of the sciences by three, and manual 
training, a foreign language, and grammar by one each. There 
were two who named mathematics as the preferred study (one 
of these, H-16, not a true homosexual). The most liked studies 
are reading, history, and literature. 

Five of the group are decidedly musical and three have been 
actors. 

More of the subjects were reared as Catholics than as Prot- 
estants, but the group is too small to lend much if any signifi- 
cance to this fact. It is interesting that two of the subjects were 



CASE STUDIES OF HOMOSEXUAL MALES 



319 



trained for the priesthood. One wonders whether they were 
encouraged to do this because they were seen to lack heterosexual 
interests, or because the mothers were so attached to them as to 
want (subconsciously) to make sure that no other woman should 
have them. 

Seven were effeminate in voice and five in manner or gait. 
Several were of the type that seems especially to attract the 
active homosexual male (petit, plump, youthful appearing, etc.). 

About half the subjects seem to have taken to homosexual 
practices more or less spontaneously; but there were four or five 
cases of downright seduction. 

Nine of the subjects say that they have never had heterosexual 
intercourse. Six have copulated with women occasionally but 
prefer men, and two have maintained regular heterosexual 
relationships over long periods (one case, no report). The mean 
M-F and "I" scores of these three groups are as follows: 





Group 1 
N = 9 


Group 2 

N = 6 


Group 3 
N = 2 


Mean M-F score. 


- 50 


- 27 


+ 17 


Mean "I" score .. 


+454 


+ 527 


+52 



We note here a striking progression toward less feminine score 
with increased tendency to heterosexuality. The difference in 
mean "I" score between groups 1 and 2 is not statistically 
significant, but the "I" scores of both men in group 3 are low for 
inversion. 

Six of the group either deny the practice of fellatio or say that 
they greatly dislike it, as compared with three who admit that 
they prefer it to pederasty. 

Several of the group give evidence of being neurotic, psy- 
chasthenic, or otherwise emotionally unstable, and in several cases 
there appears to be considerable evidence of instability in one 
or both parents. Such conditions are probably connected with 
homosexuality both as cause and effect. 

If the case-history data supplied by these individuals can be 
accepted as anywhere near the truth, the psycho-social formula for 



320 SEX AND PERSONALITY 

developing homosexuality in boys would seem to run somewhat 
as follows: too demonstrative affection from an excessively 
emotional mother, especially in the case of a first, last, or only 
child; a father who is unsympathetic, autocratic, brutal, much 
away from home, or deceased; treatment of the child as a girl, 
coupled with lack of encouragement or opportunity to asso- 
ciate with boys and to take part in the rougher masculine 
activities; overemphasis of neatness, niceness, and spirituality; 
lack of vigilance against the danger of seduction by older homo- 
sexual males. The formula, of course, does not always work. 
Doubtless many children who grow up in an environment of the 
kind just described become nevertheless heterosexual; possibly 
a majority do. In some of these cases the heterosexual adjust- 
ment is made only with difficulty; the man may have little 
interest in sex, he may select a wife much older than himself (a 
mother surrogate), or if he marries a younger woman she may 
find it impossible to win first place in his affections. 



CHAPTER XIV 

CASE STUDIES : MASCULINE AND FEMININE TYPES OF 
DELINQUENT GIRLS 1 

The M-F test was given to all the inmates of the Ventura 
(California) Home for Girls, 54 in number. The mean score 
was 88; S.D. of the distribution, 42. The mean is approxi- 
mately the same as that of high-school girls (85.5), and only 
10 points less feminine than that of eighth-grade girls. The 
group as a whole therefore rates as one of the most feminine we 
have tested. The range, however, is very great: from +103 to 
166. The case studies reported in this chapter include two 
with highly masculine scores, five with scores in the moderately 
masculine range (12 to 37), and seven whose scores were 
very feminine (166 to 132). Case studies of such widely 
contrasting groups serve to bring into relief both the value 
and the limitations of the M-F test as a psychometric tool. The 
seven most masculine girls of the series will be designated as the M 
series, the seven most feminine as the F series. The "M" cases 
follow in order of masculinity of score. 

M-l 

Exercise 1234567 Total 

M-F score 0-30 +21 +86 -3 +2 +103 

M-l is 17 years old. She is of very superior intelligence, as she 
was in the ninth grade when she left school at the age of 13. At 
16 her IQ on the Stanford-Binet test was 118, and as she passed 
all but one test in the scale it is probable that her true IQ was 
not far from 130. Her educational age on the Stanford Achieve- 
ment Test is slightly better than 18 years, which agrees closely 
with the intelligence score. 

1 The data for the case studies of this chapter were assembled by Dr. Maud 
A. Merrill with the cooperation of the officers of the Ventura Home for Girls, 
Ventura, California (a state institution for delinquents). The summaries here 
reported were prepared by Dr. Merrill and Dr. Terman. 

321 



322 SEX AND PERSONALITY 

The father of M-l is 45 years old. He works in a steel factory, 
but was formerly a mining engineer. The mother, who is 
39, did general housework before marriage. M-l is the second 
of nine children, all but one of whom (the fifth) are girls. 

M-l was never in sympathy with either of her parents, though 
she liked her father better than her mother. She never wanted 
to do the things the rest of the family did and usually disagreed 
with their ideas. The father and mother are hard-working 
people, American born of Slavic peasant stock. Their sole 
ambition has been to work hard and get ahead. M-l, too, was 
expected to work hard and to care for nothing else, and her out- 
breaks against the tyranny of dull routine seem to have been met 
only with repressive measures. She did not especially care 
for her sisters. They liked to sew and cook and she did not, and 
her older sister was "always going to dances." 

As a child M-l played with dolls, but preferred active games 
and outdoor activities. She played basketball and especially 
liked to ride horseback until tired out. She also liked to take the 
children to the basement and "make them do stunts jumping, 
etc." She wants and expects to be an aviator. Hates sewing 
and fancywork and cannot make her own dothes. She likes to 
use tools; helped her father build a garage and has built chicken 
houses by herself. Her favorite subjects were algebra and 
arithmetic; her hardest, reading. Likes the books of Zane Grey, 
"especially the adventure parts." Preferred magazines are 
Saturday Evening Post, Ladies 9 Home Journal, and the mining 
engineering magazine which her father takes. Prefers the sports 
section of the newspaper. Current news most interesting to her 
was about Byrd's expedition and Nobile's polar flight. Likes 
to go out in the rain on stormy nights. Her earliest memory is of 
"playing in the mine, taking an ax and doing like grown people." 

M-l has no intimates among her fellow inmates and few inter- 
ests in common with them. At home she belonged to a gang 
of girls who sometimes dressed up in boys' clothes and went out 
at night "to stir things up." They are said to have taken 
automobiles, picked up girls who mistook them for men, and then 
made the girls walk home. The members of the gang called 



TYPES OF DELINQUENT GIRLS 323 

themselves the "seven devils" and boasted of being able to 
"lick any boy of their size." Once they held up a dance hall, 
broke the lights, and robbed the patrons. M-l displays the 
staunchest loyalty to her former associates. 

The subject is a very attractive girl, frank and jolly. She tells 
about herself in an objective half-humorous way, but is very 
positive about her opinions. She is unemotional, self -controlled, 
of pleasant manner, and respectful in her attitude toward others. 
Her history shows that she is self-assertive, resistant to sug- 
gestion and authority, and not readily guided by advice. Her 
experiences have been predominantly masculine, and she prefers 
masculine activities and occupations. Before commitment she 
was in the habit of drinking to excess. It is not surprising that 
she claims to have always wanted to be a boy. 

M-l ran away from home, which was in a small town in Ohio, 
and ' ' hoboed ' ' her way to Los Angeles in 1 5 days, dressed in boy's 
clothes. She used a boy's name and succeeded in passing as a 
boy notwithstanding her distinctly feminine appearance. She is 
of medium height and has well-developed hips and breasts and a 
round face. Her expression is merry and she smiles readily 
when she talks. She says she left home on account of a quarrel 
with her father over a revolver, but the underlying reason seems 
to have been her rebellion against routine and her lack of sym- 
pathy with her family. Admits that she was "a bit groggy" 
with liquor when she left and that she hardly knew what she was 
doing until she got as far as St. Lotus. Soon after reaching Los 
Angeles she was arrested as a vagrant. The medical examination 
reports her as negative for venereal infection. 

Estimates of subject by the officers of the institution differ 
greatly. According to one, "she is crazy for drink; untrust- 
worthy; would do anything to achieve her ends; is interested 
only in the things that boys are interested in." According to 
another she is "very substantial, dependable under supervision." 
A third says she is "a very worth-while girl; I expect her to come 
out all right; I don't blame her for wanting to get away from the 
drudgery of her home or for refusing to return to it." This 
officer had her detailed to her own home where she (willingly) 



324 SEX AND PERSONALITY 

cooks, waits on the table, and does other housework. In her 
studies she did well in commercial arithmetic but was so unskillful 
at typing that she was removed from the dass. 

The case history of M-l, like the other histories of this series, 
was prepared by Dr. Merrill without knowledge of the results 
of the M-F test. However, Dr. Merrill rated this subject as the 
most masculine of the group and made the notation that unless 
the test showed her to be the most masculine it would be in error. 
Her score is in fact at the 73d percentile of high-school boys 
for masculinity and a full standard deviation more masculine 
than the mean of the 604 males in our adult population. It is 
beyond the 100th percentile of high-school girls. The score of 
+86 on Exercise 5 (interests) is one of the most masculine we 
have encountered. 

M-2 

Exercise 1234567 Total 

M-F score -13 -2 -6 +29 +46 -3 -2 +49 

M-2 is 17 years old and in the senior year of high school. At 16 
she earned an IQ of 112 on the Stanford-Binet test, which is 
probably equivalent to a true IQ of 120 for a subject of this age. 
Her educational age at that time on the Stanford Achievement 
Test was 16-8. The subject is therefore unquestionably well 
above the average in general intelligence. 

The father of M-2 is a mechanic, 45 years old. The mother 
died in childbirth when subject was about 13. The father 
remarried but separated from his second wife when he learned 
that she had another husband living. M-2 is the second of five 
children. The first sib is a sister, one year her senior. The 
next younger is a boy eight years her junior. There is nothing 
in her family background to account for the masculinity of her 
score. M-2 never got along pleasantly with either her father 
or her stepmother; says she never wants to see her father again. 
The stepmother is said to have had a bad influence on her 
behavior. 

In personality M-2 has a slow, plodding manner. She is 
moody, visionary, and talkative. Her first conflict with the law 
came when she was reported for truancy. A home was found 



TYPES OF DELINQUENT GIRLS 325 

for her but she again ran away and was found living with a man 
as his wife. She had a seven-months miscarriage and accused a 
second man of being the child's father. Later she lived two 
weeks with a third man. Has been under treatment for gonor- 
rhea and syphilis. 

The subject's behavior is decidedly masculine. She was for 
a time a reporter, writing sketches and illustrating them with pen 
pictures. Previous to that she had belonged to a "bicycle 
gang." After commitment to Ventura she ran away, helped 
by one of the officers, but was caught and returned. She is quite 
lacking in the feminine type of emotionality and loves adventure. 
In many respects she resembles M-l, especially in her spirit of 
independence and her qualities of leadership. Her sex involve- 
ments appear to have been incidental consequences of her 
insatiable appetite for thrills. 

The subject's M-F score is at the 100th percentile for high- 
school girls and at the 35th percentile for high-school boys. It is 
most masculine on Exercise 4 (emotional and ethical attitudes) 
and Exercise 5 (interests). 

M-3 

Exercise 1234567 Total 

M-F score -11 -2 -10 +4 +14 -6 -1 -13 

M-3 is 17 years old and in the tenth grade. No intelligence 
score is available, but the educational age of 15-6 on the Stanford 
Achievement Test indicates that her IQ is at least 100 and 
possibly higher. 

The father, a cook, deserted the family when the subject was 
6 months old. The mother was remarried, to a farmer. M-3 
has only one sibling, an older sister who is married and has one 
child. She is companionable with the sister and takes her 
troubles to her rather than to her mother. 

The subject's attitude toward her mother is unsympathetic; 
"she has such high standards for me." However, the mother is 
very indulgent and has ambitions for her daughters. The step- 
father was not good to the mother when the daughters were at 
home, and refused to give her money because she spent it for 
them. The mother was therefore compelled to go out to work 



326 SEX AND PERSONALITY 

and this increased the friction between the subject and her 
stepfather. 

As a child M-3 played mostly with boys, climbed trees, and 
preferred outdoor games. She never played house and cared 
little for dolls. "I can only remember having one doll, but 
my sister always had them." She expects to be a stenographer, 
but would like most of all to be a doctor. Favorite subjects are 
her commercial courses; the hardest, mathematics. She prefers 
to read historical books and stories of adventure; has no interest 
in love stories. Parts of newspapers liked best are the comics 
and "topics of interest, like murders." She is fond of tennis, 
baseball, volleyball, and basketball; is captain of the high-school 
basketball team. Likes to make things with tools. 

M-3 has always been a leader among girls. When in the 
Alameda Training School for Girls, where she was placed before 
her commitment to Ventura, she is reported to have developed 
"a sort of sex complex which reached a climax when the girls 
scratched her initials on their legs with pins." The subject 
herself says: "Since I have been shut up with girls I have always 
had favorites." She prefers the athletic type of girl. 

M-3 is a rather attractive girl, pleasant and docile to people in 
authority, but aggressive in dealing with girls of her own age. 
Seems to be objective and unemotional. Reports that she 
always wanted to be a boy and still does. "I've always wanted 
somebody to pal around with, but there is always a feeling of 
sex between a boy and a girl and I didn't like that. Never liked 
to go out with boys on that account. As I got older I wanted 
to be a boy just that much more." 

M-3 ran away from home (in the middle west) to find her 
father, who had deserted the family when she was an infant. 
Neither her mother nor her stepfather had any control over her 
and she was already a sex delinquent. Was sent to the Alameda 
Training School for Girls. Was paroled and placed in a family 
where she worked for her board and attended business school. 
Violated parole by going to gin parties and establishing irregular 
sex relations. Was returned to Alameda. Paroled a second 
time, left home and went to Fresno without her mother's consent. 



TYPES OF DELINQUENT GIRLS 327 

Was committed to Ventura as incorrigible. The officers at 
Ventura consider her "hard-boiled" and untrustworthy. 

The history of M-3 reveals a personality which resembles the 
masculine more than the feminine type in respect to leadership, 
aggressiveness, objectivity, interest in tools and athletics, and 
love of adventure. There is no definite history of homosexual 
practices, but the subject plays distinctly the active role in her 
relations with the other girls. Her heterosexual experiences 
have probably been incidental results of her liking for adventure. 

Only one high-school girl in 20 tests as masculine as M-3, and 
about one high-school boy in IS tests as feminine. The subject is 
accordingly about halfway between the norms for boys and girls 
of corresponding age. She is decidedly less masculine in M-F 
score than M-l or M-2, and somewhat less so in her behavior. 

M-4 

Exercise 1234567 Total 

M-F score -13 -2 +2 +3 -8 -15 -2 -35 

M-4 is 18 years old and a high-school senior. Her Stanford- 
Binet IQ is 104, probably equivalent to a true IQ of 115. The 
Stanford Achievement Test gives her an educational age of 17-2. 

The father is 46 years old. He is a mechanic and has a garage. 
The mother, who was IS years older than her husband, died when 
M-4 was 10 years old. At the time of her death she had been 
separated from her husband for two years and was a teacher. 
The father has remarried. M-4 has an older sister who is 
married and has three children. There are two half-sisters, still 
of preschool age, by her father's second marriage. The subject 
was very congenial with her mother: "The two years mother and 
I had together [after parents' separation] were the happiest in 
my life." 

After her mother's death M-4 went to live with her father, who 
had by this time remarried. She disliked her stepmother but 
did not quarrel with her. The father allowed her little freedom, 
kept her at home, and established incestuous relations with her. 
When this became known she was committed to the New York 
Training School for Girls. After parole she ran away to Cali- 
fornia where an aunt lived, but the latter refused to take her in 



328 SEX AND PERSONALITY 

or have anything to do with her. Soon afterward she was 
arrested in an apartment with a man and was brought into the 
juvenile court on the charge of vagrancy. She was found to be 
infected with gonorrhea and was committed to Ventura. 

M-4 has the dignity and poise of a woman of superior social 
level; gives the impression of being rather inscrutable and 
complete mistress of herself. She is reticent but when she 
talks she expresses herself in good English. Told somewhat 
reluctantly of the incestuous relationship with her father. 

The subject's experience has been predominantly feminine. 
As a child she preferred girls to boys, but played with both. 
Liked dolls, but was also fond of skating and outdoor games. 
Was captain of a girls' baseball team. Her favorite school 
subject was mathematics; her hardest, English. Her favorite 
books were Campfire Girls, Gene Stratton Porter's novels, and 
the Elsie books. She reads American Magazine and Good 
Housekeeping. Has never used tools. Can sew and do plain 
cooking, but does not care for fancywork. Expects to study 
medicine. 

On the M-F test, M-4 rates more masculine than six girls out of 
seven at the high-school age. In Exercise 3 (information) and 
Exercise 4 (emotional and ethical attitudes) her scores are more 
masculine than feminine. Her history strongly suggests, how- 
ever, that the subject has essentially a normal and well integrated 
personality, and that her delinquencies were the result of unfor- 
tunate circumstances. Her activities and interests have certainly 
been more feminine than the score would lead one to expect. 

M-6 

Exercise 1234567 Total 

M-F score -22 -1 -6 +17 -20 -2 -2 -36 

M-5 is IS years old and in the ninth grade. No intelligence 
score is on record, but the educational age of 16-2 on the Stanford 
Achievement Test would suggest an IQ appreciably above 100. 

Subject's father, aged 40, is an airplane machinist. The 
mother, also 40, did housework outside her home for a year after 
marriage. There are five children, of whom M-5 is the second. 



TYPES OF DELINQUENT GIRLS 329 

The first is a girl of 17, the third a boy of 10, the fourth died 
at the age of 3, and the youngest is only 2. 

The home relationships were anything but sympathetic. 
"Mother understood me better than my father, but she did not 
see things from my point of view. She would never let us 
[subject and older sister] go out; said we could not be trusted. 
So we went to dances without her permission." Regarding 
relationship to her sibs, "We didn't play, we fought." However, 
she was fond of the youngest brother and liked to take care of him. 

M-5 played with dolls until she was 6 years old. Liked to 
climb trees, play games, and go flower picking. She wanted to 
become a Girl Scout, but the mother was unwilling to let her 
attend the Scout meetings after school. Had no girl chums. 
"Only one girl lived near us and mother wouldn't let me have 
anything to do with her. But they let me go out with a man of 
20, a friend of father's." Always wanted to be a boy; "if I were 
I would be an aviator." She likes to use tools and "can make 
almost anything." Can sew, embroider, and do plain cooking. 
Asked what she expected to do when grown up, she said: "Marry; 
nearly everyone does." Favorite subject, arithmetic; hardest, 
civics. Likes to read adventure stories. Dislikes athletics. 

M-S is a gloomy and unresponsive girl, bitter, and resistant 
to advice. She justifies herself and blames her family for all her 
troubles. Appears sullen, hard, and suspicious. Was com- 
mitted on the charge of having sexual relations with several 
boys. However, she denies having had normal sexual inter- 
course. (Record: "Ruptured hymen; probably has had inter- 
course.") The house mother in charge reports that M-5 is 
continually doing things that have a sexual significance, such as 
handling the other girls. Is described by another officer as "very 
eccentric and queer." 

Only one girl out of seven of corresponding age scores so 
masculine on the M-F test as this subject. Her most masculine 
score is on Exercise 4 (emotional and ethical attitudes). It 
would be interesting to know to what extent her rebellious 
behavior is the cause and to what extent the effect of her mas- 
culine attitudes. As the older sister (not tested) was also for a 



330 SEX AND PERSONALITY 

time in a detention home, it is possible that the domestic situa- 
tion may be largely responsible for provoking the resistant and 
otherwise masculine type of personality. 

M-6 

Exercise 1234567 Total 

M-F score -IS -2 -4 +18 -28 -3 -3 -37 

M-6 is 20 years old. No intelligence-test record is available, 
but her high Stanford Achievement Test score and the fact 
that she reached the twelfth grade before she left school at 16 
(or 17) indicate that her IQ is probably above 120. 

The subject is an only child, her one sib (just older) having died 
at birth. This circumstance has probably been a major factor 
in the subject's life history. The father is a retired contractor, 
58 years old. The mother, 57, was formerly a singer; sang with 
the choral symphony of one of the largest middlewestern cities. 
M-6 is married and most of her troubles have arisen from her 
mother's inability to give her up. "Mother would give her life 
for me, but she won't give me up to my husband." 

The fact that the first child had been stillborn, and that the 
mother was 37 at the birth of M-6, doubtless accentuated the 
only-child situation. From infancy the child was pampered 
and adored by her mother. The latter has twice deliberately 
lured her daughter away from her husband; writes her every 
day the most effusive letters protesting her affection for her 
" darling child." Subject is torn between husband and mother: 
"I can't live with either of them, and I can't live without them." 
She resents any show of authority by either. The father seems 
to play almost no part in this conflict. 

As a child M-6 liked sports and games better than the activities 
more typical of girls. Her mother provided her with innumerable 
dolls, but the only one she cared for was an old rag doll. She 
liked to play with other children, preferably boys. She likes 
art, music, and dramatics; is in fact very musical and has always 
been a leader in dramatics. On the whole, she prefers the stage 
to music. She also likes business and has held several office 
positions. She is good at mechanical drawing; was the only 
girl in a class of boys in this subject and was outstanding. She 



TYPES OF DELINQUENT GIRLS 331 

likes to do things that require fine coordinations and meticulous 
care. Favorite subjects, English and Spanish; hardest, history 
and geography. Excelled in swimming and hunting, though 
her mother thought it wasn't ladylike for her to go hunting. The 
subject differs markedly from her fellow inmates in the superior 
quality of her interests, training, and background. 

M-6 is vivid, beautiful, charming, and merry; the kind who 
could hold her own in any social group. She is whimsically 
objective, interested in everything, happy, dominated by her 
affections, quick to resent authority, loyal and fearless. Never 
wanted to be a boy; "I could always play with the boys anyway, 
and they liked to play with me better than with other boys." 

M-6 was committed for forgery. The mother, after luring her 
away from her husband, encouraged her to accept the attentions 
of an older man. For a time she was infatuated with this man 
and, although he was not divorced, lived with him as his wife. 
He taught her to cash forged checks at department stores after 
banking hours, and in a spirit of lawless adventure and revolt 
she entered into this new game. She is not popular with the 
officers of the institution because they cannot break her spirit. 
She is never insolent, but she is never cowed. 

M-6 was 17 when she eloped and married. All went well as 
long as she and her husband could live with her mother, but 
when he took a position in another city the trouble began. M-6, 
much against her mother's wishes, also got a position. Her 
mother kept trying to get her to come home, which she did one 
day after a quarrel with her husband. Later reunited with her 
husband, but returned home to get her household goods and 
was persuaded by her mother to stay. The mother's next 
stratagem was to encourage her to accept the attentions of the 
other man in order to keep her from her husband. When 
M-6 was arrested for forgery this man behaved in a most cowardly 
and disloyal way, yet she refused to turn state's evidence. 

From the case history it was expected that this subject would 
score somewhat masculine, but within the feminine range. Such 
was the case. One might say that she is a man's woman, par 
excellence. 



332 SEX AND PERSONALITY 

M-7 

Exercise 1234567 Total 

M-F score -14 -3 -7 +30 -38 -7 +2 -37 

M-7 is 17 years old, has a Stanford-Binet IQ of 90, and rates a 
little above 13 years educationally on the Stanford Achievement 
Test. She is at best not above average in mental ability. 

The parents were divorced when M-7 was 5 years old and the 
mother has remarried. Her stepfather is a carpenter and has 
served a term in prison. Her mother was a telephone operator 
before marriage. There are two half-sibs, both boys. Subject 
is congenial with her mother but has never felt free to talk to her 
about her affairs. 

As a child M-7 never played with dolls and was very fond of 
outdoor activities. She prefers boys to girls as friends. ' ' Always 
as a little girl I liked to play with boys better than with girls; 
you can tell them more." However, she did not want to be a 
boy. At present gives the impression of being poised, self- 
confident, and unemotional, which is in line with her very 
masculine score on Exercise 4, emotional and ethical attitudes. 

The subject was only 12 years old when she was first brought 
into the juvenile court. She was then under treatment for 
gonorrhea. After a short time in a training school for girls she 
was released and worked as a clerk while living at home. Because 
the mother objected to her associates she ran away, was appre- 
hended, and tried to commit suicide. It was after this that she 
was committed to Ventura. 

This case presents no very unusual features from the point of 
view of M-F characteristics. Although only about one girl out 
of seven at her age tests as masculine, she is well within the 
normal feminine range. Her life history indicates that her 
experiences and interests have presented a mixture of masculine 
and feminine trends. Her favorite school subjects were civics 
and hygiene; her hardest, arithmetic and English. She likes 
dancing, romantic reading and movies, cooking and sewing. Has 
ushered in a movie theater and would like to be an interior 
decorator. Has played on a basketball team. One wonders 
whether the unemotional and somewhat "hard-boiled" attitudes 



TYPES OF DELINQUENT GIRLS 333 

reflected in her responses in Exercise 4 are the cause or effect 
of her delinquent tendencies. 

Before proceeding to the F group of female delinquents a few 
summary observations may be made regarding the M group of 
the series. 

1. All the subjects in the group are above the 85 th percentile of 
high-school girls for masculinity; two are above the 100th 
percentile. 

2. The group as a whole rates high in intelligence. Four have 
estimated IQ's between 115 and 130, the highest belonging to the 
most masculine-scoring subject. Two are at or just above 100, 
and only one is below. 

3 . Of the fathers of these seven girls, five belong to occupations 
of the engineering or mechanical type. The father of the most 
masculine-scoring and most masculine-behaving girl has been 
a mining engineer (this occupation is the most masculine in our 
tested populations). Three other fathers are mechanics and one 
is a building contractor. 

4. The group as a whole has shown far more than the usual 
amount of interest in masculine activities. This is particularly 
true of the three whose M-F scores were most masculine, and 
most of all true of the subject who ranked highest in M-F score. 
The four least masculine showed a distinctly greater mixture of 
masculine and feminine interests. Five of the seven cared little 
or not at all for doll play. Athletics or other outdoor activities 
figure prominently in the interests of six. Two were members 
of gangs. Three liked to work with tools. Reading preferences 
of three were chiefly masculine. Not one was particularly fond of 
cooking, sewing, or fancywork. Three have always wanted to be 
boys. Their vocational preferences are for the most part masculine. 

5. In general the subjects of the M group are of aggressive, 
independent spirit; they are self-confident, well poised, and 
lacking in feminine emotionality. One at least (M-4) is more or 
less psychopathic. 

6. Exercises 4 and 5 contribute most heavily to the masculine 
scores of the M group, indicating unemotionality of response and 
a marked tendency to masculine interests. 



334 SEX AND PERSONALITY 

In going through each of the case histories of the F series which 
follows, the reader should keep in mind so far as possible the 
composite picture of the M group in order to note both contrasts 
and resemblances. The F cases are presented as ranked by the 
M-F score from most to least feminine. 

F-l 

Exercise 1234567 Total 

M-F score -23 -2 -22 -56 -46 -15 -2 -166 

F-l is 18 years old. Both her Stanford-Binet and Stanford 
Achievement scores indicate that she is of about average 
mental ability. She is the only child of American-born parents 
of Spanish descent. The subject's illness prevented the taking 
of a case history, but the following information is available from 
the institution records. 

F-l was brought to the San Francisco Detention Home for 
truancy at the age of IS and was found to have both gonorrhea 
and syphilis. She admitted sexual relations with sailors. Was 
sent to Convent of Good Shepherd, but was soon paroled and 
continued her former associations. Was again sent to the 
convent and again paroled. Then her mother died and as the 
father was unable to control her she was committed to Ventura. 

F-l is said by one of the institution officers to be oversexed. 
She wants to become a stenographer, is fond of music, public 
dances, tennis, and roller skating. Is of the quiet type and reads 
a good deal. Feminine interests seem to have predominated, but 
the available information is scanty. In the absence of full 
personal history data the case is of interest chiefly as an illustra- 
tion of one of the extremes of the wide M-F range which one 
finds among delinquent girls of the late teens. Only one high- 
school girl in SO tests as feminine as this subject, her score on 
Exercise 4 (emotional and ethical attitudes) being especially 
feminine. 

F-2 

Exercise 1234567 Total 

M-F score -27 -1 -2 -42 -78 -1 -2 -153 

F-2, aged 17, is considerably above average in mental ability. 
Her found IQ of 111 at age IS suggests a true IQ in the neighbor- 



TYPES OP DELINQUENT GIRLS 335 

hood of 120, and her achievement scores are correspondingly 
high. 

The mother, now 36, married the father, a preacher's son, when 
she was 19. F-2 accuses her father of cruelty to her mother and 
of infecting her with syphilis. Mother deserted husband when 
subject was 14 months old and after teaching school for several 
years married a fairly well-to-do man whom she did not love 
and lived with only three years. Subject's only sib, her twin, 
was stillborn. F-2 is very devoted to her mother and takes a 
protective attitude toward her. 

As a child F-2 loved to play with dolls. " I still do. Anybody 
can laugh that wants to. I still have a large doll." Was 
prevented by her mother from associating with other children, 
so stayed at home, read a great deal, worked in the garden, and 
rode horseback. "Most of the things I did were solitary." Her 
favorite subject was English; her hardest, arithmetic. Favorite 
author is Gene Stratton Porter. Reads Pictorial Review, 
Delineator, Ladies 9 Home Journal, Red Book, American Magazine, 
"the sensational parts of the newspaper and the woman's 
page." Is extremely fond of dancing, pets, embroidering, and 
sewing. "I love to sew; it is an inherited instinct." Expects 
to be a stenographer, but would like to be a novelist. Writes 
poetry. 

F-2 is a very unstable girl. She is emotional, but her emotions 
are shallow. Is uncontrolled, self-centered, selfish, and obsessed 
with sex interests. Scored highly psychoneurotic on the Wood- 
worth-Cady personal inventory (-2.2 S.D.), at the 100th 
percentile for instability on the Laird B2, and showed emotional 
disturbance on the Wells association test, especially on the 
Jung words that tap sexual complexes. Is described by the 
institution officers as queer and completely solitary; with one 
accord they say she is unstable and will probably become insane. 

F-2 first came into the juvenile court because of her illicit 
sexual relations with several men, including two university 
students. She was placed in a training school for girls, but was 
found so unstable that she was returned to the court as an unfit 
subject for training. She was then committed to Ventura. 



336 SEX AND PERSONALITY 

Significant facts in this case history are (1) the strong mother 
attachment, probably resulting from the mother's unhappy 
marital experience; (2) the lack of opportunity to associate with 
other children; and (3) emotional instability with persistence of 
infantile egoism. In the M-F test her most highly feminine 
scores were made on Exercise 1 (word association) and Exercise 5 
(interests). 

F-3 

Exercise 1234567 Total 

M-F score -32 -2 -21 -6 -74 -9 -3 -147 

F-3 is 19 years old. Her score on the Stanford Achievement 
Test suggests an IQ of 100 to 110. Reached the junior year of 
high school and became a stenographer. 

At the age of 14 F-3 lost her mother; two years later her father, 
who keeps a fruit stand, remarried. She has one younger 
brother and two younger sisters. 

The subject was committed for forgery. The facts are that 
she was the trusted stenographer of a businessman in a small 
California city, fell in love with a young man, allowed him to 
drive her to Los Angeles, and when he was arrested and fined 
$50.00 for speeding she signed her employer's name to a check 
in payment of the fine. 

Nothing in her childhood history is noted that is not typically 
feminine. She is very sensitive, fine spirited, sweet, and gentle. 
She is extremely remorseful and is regarded by the officers of the 
institution as essentially trustworthy. She paints well, does 
needlework, and loves to arrange flowers. Has none of the 
aggressiveness characteristic of so many delinquent girls. Her 
score is highly feminine on Exercise 1 (word association), Exercise 
3 (information), and Exercise 5 (interests). 

F-4 

Exercise 1234567 Total 

M-F score -23 -1 -3 -49 -68 -2 -1 -147 

F-4 is 18 years old and in the ninth grade. Her IQ, as esti- 
mated from her Stanford Achievement score, is about 90. She 
is the eighth of nine children in a Spanish family. Her father, 



TYPES OF DELINQUENT GIRLS 337 

owner of a store, died when the subject was 5 years old and her 
mother when she was 13 . A brother, next older, was at the time 
of this study a featherweight fighter of some note. 

F^r is devoted to her family; her eyes shine when she talks of 
them. The pugilist brother is her hero. She had a strict 
upbringing, as is customary in Spanish homes, and was not 
allowed to play much with children outside the family. Loved 
to play with dolls. Her favorite subject was arithmetic, English 
her hardest. She is fond of dancing, and likes to sew and do 
fancywork. Expects to be a seamstress. Likes to read stories 
of ancient history and romances. Favorite book, Ramona. 
Reads Pictorial Review, Woman's Home Companion, Cosmo- 
politan, and True Romances. Loves movies. Valentino was 
her favorite actor, later Novarro. Prefers to be with girls rather 
than with boys. 

F-4 is gentle, amiable, and wants to be loved. She is all 
sweetness, softness, and charm, has the social and manual 
interests of her sex, is docile and would do anything to please. 
She was committed on the charge of "leading a lewd and immoral 
life." Subject says only one man was involved and that "he 
was all right, only he was married." Shows no resentment or 
bitterness over her commitment. Is a devout Catholic. After 
taking down her case history Dr. Merrill appended the note: 
"This subject should test feminine if anyone ever did." Her 
score in fact places her at the Sth percentile of high-school-junior 
girls for masculinity (95 th percentile for femininity). The 
feminine score was earned chiefly on Exercise 1 (word associa- 
tion), Exercise 4 (emotional and ethical attitudes), and Exercise 5 
(interests). 

F-6 

Exercise 1234567 Total 

M-F score -31 -1 -8 -22 -60 -20 -2 -144 

F-5 is 18 years old and in the tenth school grade. Her IQ, 
estimated from Stanford Achievement Test, is probably in the 
neighborhood of 110. 

Subject's father, a locomotive engineer, died several years ago 
and the mother remarried. The only sib is a brother, aged 24. 



338 SEX AND PERSONALITY 

F-5 is sympathetic toward both mother and stepfather. " Could 
always talk over problems with my mother. She let me have my 
way too much." Of her stepfather she says, "He always treated 
me fairly and was rather inclined to be overindulgent." 

As a child F-S played with dolls until about 11 years old and 
made clothes for them. She played games, too, mostly with 
girls. Her favorite subject, arithmetic; hardest, composition. 
Favorite books, Grace Richmond "and that type, Red Pepper 
Burns, and love stories." Reads little in the newspapers but the 
society page. Likes fancywork and makes her own clothes; also 
plays baseball at Ventura. Likes to go out a great deal, prefer- 
ably with older companions. When only 14 was permitted to 
work in restaurants and in private homes with little supervision 
and formed a number of undesirable associations. 

F-S never wanted to be a boy; says she does not envy them 
even though she recognizes that they have better business 
opportunities and can "get by" with things girls can't. In 
personality she is amiable. A little plump, languorous in speech 
and movement, but mentally alert and willing to respond; some- 
what of the Mae West type. Would probably rate high in 
physical attractiveness to the opposite sex. The officers describe 
her as a dependable girl, with wholesome interests, poetic, and 
"probably strongly sexed." Her score is especially feminine on 
Exercises 1,5, and 6. 

Subject was arrested as an inmate of a "disorderly house" 
which was raided by the police. She denies having been promis- 
cuous but the records show that she has had a syphilitic infection. 

F-6 

Exercise 1234567 Total 

M-F score -25 -2 -4 -20 -62 -16 -5 -134 

F-6 is 18 years old. She is still in the eighth grade, has an edu- 
cational age of only 14-6, and an IQ of 92. 

The subject's parents have been divorced and have both 
remarried. The stepfather is a traveling auditor, 66 years old; 
the mother is 40. Subject is in effect an only child; there is one 
half-sib, less than 2 years old. Got along well with her mother 



TYPES OF DELINQUENT GIRLS 339 

and stepfather ("he was as good as gold to me"), and was con- 
genial with the two grown daughters of her stepfather. Records 
indicate that her mother is immoral and alcoholic. F-6 was 
for this reason sent to California to live with her father. She 
was found to be pregnant when she arrived. An uncle became 
infatuated with her and at her father's request she was made a 
ward of the juvenile court and placed at the Business Girls' Club 
in Los Angeles. She did not do well there and was sent to the 
Detention Home, from which she ran away. 

As a child F-6 liked to play with dolls and "had most fun 
dressing up in long dresses." Does not like to sew or embroider. 
Has used tools in making doll furniture, has been captain of a 
baseball team, and likes athletics. "I would rather swim than 
eat." Likes to dance. Favorite subject, arithmetic; hardest, 
history. Likes to read; favorite author is Gene Stratton Porter. 
Used to like girls her own age but gets along better with boys. 
Likes older boys and enjoys skating, swimming, and dancing with 
them. Always wanted to be a boy and still does. In personality 
is rather colorless, quiet, and unobtrusive, with a certain dignity 
of manner. Is given to crushes on the other girls of the institu- 
tion and is the object of their crushes. 

F-6 presents a combination of masculine and feminine behavior 
traits that would lead one to expect her to make a less feminine 
score than she does on the M-F test. The total score is at the 
16th percentile of high-school girls for masculinity. 

F-7 

Exercise 1234567 Total 

M-F score -21 -3 +1 -25 -80 -1 -1 -130 

F-7 is 17 years old and in the fifth grade, though her score on 
the Stanford Achievement Test shows seventh-grade mastery. 
Has had no intelligence test but the IQ estimated from achieve- 
ment score is probably below 90. 

Subject has a sister three years older and a brother two years 
younger; is fond of both. The father is a sheet-metal worker, 
aged 43 . The mother, formerly a housemaid, is 39. Her parents 
were good to her, but she considered them too strict and ran away. 



340 SEX AND PERSONALITY 

As a child F-7 played with dolls. Still likes them. Also liked 
to play school, but was not with other children very much. 
Likes cooking and nursing; "would rather be a nurse than any- 
thing." Can sew and do fancywork. Favorite school subject, 
arithmetic; hardest, civics. Favorite books, Zane Grey's. In 
the newspapers reads recipes, advice to young girls, and the 
woman's page. Never wanted to be a boy. 

In personality F-7 is suave, smooth, and colorless. Always 
talks for effect. Is described by officers as "too nice;, quiet and 
inoffensive, but a talebearer; stubborn, conceited, smug." She 
was committed for incorrigibility. She had run away from home 
a number of times and in one case was not found for a year. 
With an older woman was consorting with sailors and had 
contracted gonorrhea and syphilis. 

The experience of F-7, apart from her truancy, has been pre- 
dominantly feminine. Her M-F score is especially feminine on 
Exercises 5 and 1, and masculine on Exercise 3. 

The F group of this series presents a very different picture 
from the M group. Among the outstanding differences are the 
following: 

1. The members of the group range in M-F score from about 
the 2nd to the 20th percentile of high-school girls for masculinity 
(M group 85th to above 100th percentile). 

2. The average intelligence is lower. The group contains one 
subject of about 120 IQ, three between 100 and 110, and three 
around 90. 

3. Of the fathers of the seven only two were in mechanical 
occupations (M group, five). 

4. The interests and activities of the group have been pre- 
dominantly feminine. Only F-6 presents a partial exception. 
Doll play was engaged in by nearly all and still is by two. Only 
one showed much interest in tools, athletics, or other vigorous 
outdoor activities. None was a member of a gang. The reading 
preferences of at least five are markedly feminine, of none 
markedly masculine. Nearly all were fond of sewing and cook- 
ing. The occupational ambitions of nearly all are distinctly 
feminine. 



TYPES OF DELINQUENT GIRLS 341 

5. The F subjects, judged by their life histories, lack the 
aggressiveness, rebelliousness, independence, and self-confidence 
of the M group. They are more gentle, docile, quiet, and color- 
less in personality. Although three of the group have been 
truants, as compared with five of the M group, the truancy of the 
two groups differs greatly in character. The truant F subjects 
simply left home to associate with men in the same or a neighbor- 
ing city; the M truant is more likely to rebel against parental 
authority and to go to distant parts of the country. There was 
a somewhat greater amount of sexual promiscuity in the F 
group. Except for the sexual delinquency prevalent in both 
groups, one could say that the interests and activities of the M 
and F subjects show a very small amount of overlap. Even sex 
delinquency has not always the same significance in the two 
groups; with M subjects it is likely to be an incidental, with F 
subjects a more central, factor. 

6. The exercises of the M-F test on which the F group scored 
most feminine are 1, 4, 5, and 6. Two of these, 4 and 5, were 
most often highly masculine in the M group. So far as the test 
is concerned, the most characteristic contrasts between the two 
groups are therefore in (a) emotional and ethical attitudes, and 
(V) interests. 



CHAPTER XV 
CASE NOTES: MISCELLANEOUS 

The case descriptions we have brought together in this chapter 
represent a considerable variety. They have been selected with 
a view to giving the reader an idea of the limitations as well as the 
positive value of M-F scores in clinical practice. We have there- 
fore included a number of cases which run counter to expectations 
with respect to correlation between M-F score and overt behavior 
or interests. A test of the M-F type has, we are convinced, great 
potential usefulness in the study of personality in its relation 
to social and sexual adjustment, but it is only one of many 
methods of approach to the study of these problems and in the 
present stage of its development it raises many questions which 
it does not answer. 

X-l 

X-l is our most masculine- testing woman. She is a university 
professor, not far from 40 years of age, and is an able teacher and 
researcher. 

Exercise 1234567 Total 

M-F score: +9 -2 +5 +69 +46 +4 +1 +132 

The total score is at the 85 th percentile for college men. She rates 
on the masculine side in every exercise except the unreliable No. 2. 
The M-F score of her husband (+89) although at the 60th 
percentile for college men, is less masculine than her own by more 
than one S.D. His scores on the separate exercises are as follows: 

Exercise 1234567 Total 

M-F score: +5 +1 +10 +43 +28 +3 -1 +89 

It will be noted that the wife is especially more masculine on 

342 



CASE NOTES: MISCELLANEOUS 343 

Exercise 4 (emotional and ethical response) and on Exercise 5 
(interests). 

In view of the extraordinarily masculine score of this woman, 
and the fact that she seems to have made a normal heterosexual 
adjustment, the following communication from her is of surpass- 
ing interest: 

I might give a brief case-study of myself, inasmuch as I would classify 
as a woman brought up by men. My mother died soon after my birth 
and my education was almost wholly in the hands of my father, grand- 
father, numerous uncles, and brothers. I did not attend school until I 
was IS; up until that time I was trained at home by either my father, 
my grandfather, or a tutor, who was also a man. I dressed in boy's 
clothes until I was 13, except for such few dressed-up occasions as might 
warrant a more feminine costume. Occasionally, an aunt or other woman 
visited us and then I had to wear girl's clothes which probably started 
me on my decided antagonism toward women. My amusements were 
exclusively masculine. I do not remember owning dolls or other girls' 
toys. I played baseball, football, hockey, went swimming in the "old 
swimmin' hole," climbed every tree, house, and barn in the neighborhood, 
and so on. My playmates were all boys; the only girls I recall were girls 
I was afraid of. My chief amusement consisted in seeing how far I 
could travel from roof to roof and tree to tree without touching the 
ground I think my greatest achievement was traversing some ten blocks 
before too long a distance between trees, houses, or fences caused me to 
descend. Another amusement consisted in running along beside some 
friend who had a bicycle, holding on to the springs under the seat; by 
the time I was 10 1 could run the five miles around a near-by lake without 
stopping and with the greatest enjoyment. At the age of 15 my father 
suddenly discovered that he had a first-class boy on his hands, and that 
I was unable to get along with girls, so he sent me to a girls' boarding 
school for four years and then to a girls' college. The net result seemed 
to be that I disliked most girls and women worse than ever, fell desperately 
in love with a few, and was everlastingly being ruled out of games because 
I was "too rough." One summer I played professional baseball, my 
position being catching behind the bat; I still play baseball here on our 
faculty team in intramural ball. I also play field hockey, basketball, and 
water polo, but have to be most terribly careful not to hurt someone! My 
friends are still all men, I don't like girl students in my classes, and I 
haven't the vaguest idea what makes most women react the way they do; 
in the company of girls and women I remain utterly silent because I 
have learned that if I talk I will either bore the poor dears or else shock 
them. 



344 SEX AND PERSONALITY 

X-2 

X-2 is a woman psychologist of brilliant intellect, age between 
34 and 40. 

Exercise 1234567 Total 

M-F score: -8 -2 +9 +67 -2 +15 +79 

The total score is at the SSth percentile for college men. The 
score on Exercise 4 (emotional and ethical response) is among 
the most masculine we have ever had from a woman and is at the 
95th percentile for college men. The subject is unmarried but 
is known to have normal interest in men. She is objective and 
entirely lacking in feminine coyness. She works better with 
male than with female colleagues. Both this case and the 
preceding one are impressive illustrations of the fact that a high 
masculine score in a woman is not incompatible with hetero- 
sexual interests and adjustment. 

X-3 

X-3 is a 23 -year-old woman who was arrested for masquerading 
in men's clothes. She was tested in jail. The jail physician 
(female) contributed some of the items of information in the fol- 
lowing case summary. 

Exercise 1234567 Total 

M-F score: -24 -36 +74 -3 +11 

Stenquist mechanical-aptitude score near to mean for high- 
school girls. 

X-3 attended high school for three years and made good 
marks. Her father has been dead a year. Her mother is 
inclined to be suspicious and has a disagreeable temper when 
angered. The subject has five sisters and one brother. The 
parents of X-3 wanted a boy and have always treated her as one. 
She is called "Harry" by the rest of the family. When she was 
8 years old she began wearing boys' jeans at her mother's wish, 
and her father taught her boxing. However, she was sent to a 
girls' private school, where the girls all called her a tomboy. 
She never cared for dolls and dislikes housework and babies. In 



CASE NOTES: MISCELLANEOUS 345 

school plays she always took boys' parts and her teachers called 
her "Tommy." She learned to play the saxophone. 

Later she found it fun to dress as a boy and go out with chorus 
girls and to rouse their emotions by kissing them as boys would. 
She has no feeling for boys but is sexually excited by girls, 
especially blondes. Says it would be easier to steal than to be a 
prostitute. 

Shoulders 35^ in., hips 34 in.; breasts well developed and 
distribution of pubic hair is feminine; large amount of hair on 
her legs; voice contralto; arms and neck rather muscular; stands 
with feet apart; avoids color in dress. 

In view of the early environmental conditioning it is not 
surprising that X-3 is more masculine in the M-F test than 95 
out of 100 women of her age and education. It is interesting to 
note that her score is more than normally feminine on Exercises 
3 and 4, while on Exercise 5 it is one of the most masculine we 
have encountered among women. 

X-4 

X-4 is an artist, aged 27. She was the victim of a homicidal 
attack by a female " lover" who had learned of her marriage to a 
man. X-4 had played the passive role in this homosexual 
attachment, which was of long standing. 

Exercise 1234567 Total 

M-F score: -20 -1 -25 +17 -36 -11 +1 -75 

This score is if anything more feminine than that of the generality 
of women of the subject's intellectual and educational class; 
possibly not unusual for the passive member of female homo- 
sexual partnerships. The subject had an earlier marriage which 
lasted four years. Her self-ratings on interests were normally 
feminine. Her homosexual mate who assaulted her could not, 
unfortunately, be persuaded to take the test. 

X-6 and X-6 

X-5 and X-6 are members of a homosexual pair. Both are 
coflege graduates, teachers, and in their thirties. X-5 the active, 



346 SEX AND PERSONALITY 

X-6 the passive member of the union. The score of X-6 is near 
the median for college women; that of X-5 is near the 75 th 
percentile for college men. 

Exercise 1234567 Total 

X-5 4-2-1+5 +36 +31 +6 +2 +81 

X-6 -31 -2 -11 +22 -40 -3 -2 -67 

X-S is tall and slender with boyish build. Her clothes are 
rather severe and lack the feminine touch. She is quiet and 
efficient and is well liked by her acquaintances, especially by 
men, who admire her for her intelligence and dependability. 
She has marked mechanical interests. Has had no affairs with 
men. 

X-6 is small and of feminine build. Her father was brutal, and 
she left home at a tender age. After graduation from college she 
married an effeminate man who left her because he disliked 
sexual intercourse. The subject has since lived with X-5, who 
looks out for her in the manner of a devoted husband. Outside 
her work, which she does expertly, she exemplifies the helpless 
type of femininity. There is evidence that but for her 
unfortunate marriage she would have been entirely heterosexual. 

Our data on homosexual women are too limited to warrant 
generalizations, but we are inclined to think that the M-F 
scores of these two women are probably fairly typical of the 
active and passive types. Thus far we have found no active 
homosexual female with a normally feminine score, and no 
passive with a score that was highly masculine. 1 

X-7 and X-8 

X-7 is a Canadian teacher, 34 years old. She is described by a 
psychologist acquaintance as "a sublimated homosexual given 
to intense heroine worship, and the passive rather than the 
active member of these attachments." 

Exercise 1234567 Total 

M-F score: -15 -1 -16 +10 -28 -16 -1 -67 

1 See appendix for data on other homosexual women. 



CASE NOTES: MISCELLANEOUS 347 

X-8 is a teacher of physical training who has greatly annoyed 
some of her friends by her homosexual advances, which included 
impassioned love letters and attempted physical caresses. 

Exercise 1234567 Total 

M-F score: -20 -4 +7 +90 +4 +1 +78 

These two cases present an interesting contrast, one of them 
scoring on the M-F test near the average of women and the other 
near the average of men. It can hardly be accidental that their 
homosexual tendencies are in opposite directions. The score 
of X-8 on Exercise 5 is above the 95th percentile for men. 

X-9 

X-9 is a high-school girl of 17. She was given the M-F test as 
a result of the high-school principal's suspicion that she was a 
homosexual. She had developed very intense " crushes " on 
other girls and had carried her attentions to the point of making 
them decidedly objectionable. 

Exercise 1234567 Total 

M-F score: -18 -11 -12 +16 +10 -1 -17 

According to the test, only 1 high-school girl in 20 is as masculine 
as X-9. She is much interested in travel, outdoor sports, science, 
and pets, and has little or no interest in religion, music, or 
domestic arts. These interests, except for the positive interest 
in pets, are typical of the masculine type of girl. 

X-10 

X-10 is a psychologist in her late twenties. 

Exercise 1234567 Total 

M-F score: -13 -3 -5 +46 -38 +2 -1 -12 

This score is near the 92d percentile of college women for mascu- 
linity ; about 1 college woman in 12 is as masculine. The score on 
Exercise 4 (emotional and ethical attitudes) is especially mascu- 
line, which seems not uncommon among women psychologists. 
The subject is married and well adjusted. 



348 SEX AND PERSONALITY 

X-ll and X-12 

X-ll is a male, 24 years old, who was arrested for attempted 
rape on a girl of 5 years. His M-F score of +116 is at the 75th 
percentile for high-school males. 

X-12, female, was tested in a university hospital because of 
suspected nymphomania. The suspicion was based on numerous 
improper advances to internes and medical students, with 
extreme disregard for the presence of observers. Her highly 
feminine M-F score of 123 (25th percentile for masculinity) is 
only partly accounted for by her low mental age of 13 years. 

Nothing is known as to whether highly masculine and highly 
feminine scores are, respectively, characteristic of males and 
females given to unrestrained sexuality. 

X-13 

X-13, a boy, was 13 years old when brought to the university 
for examination because a school psychologist had pronounced 
him an invert. 

Exercise 1234567 Total 

M-F score: +6 -9 -11 +22 -15 -1 -8 

The mother of X-13, a teacher in an extremely isolated rural 
district of California, furnished the following history. She was 
40 and the father was 46 when the child was born. Subject has 
two sisters, aged 22 and 19, and a brother of 15. The brothers 
have nothing in common. Between the age of 1 and 2 years 
X-13 did not seem to be developing properly and was given thy- 
roid extract. Improved. Entered school at 6; had "hysterics" 
the first day. Has never become socialized as far as boys are 
concerned, but was popular with girls to the age of 7. At recess 
he always stayed inside unless forced to go out. 

The subject was only 7 when he came to his mother sobbing: 
"Mother, I'm different, I'm different." She tried to assure him 
he was not. Soon afterward he began to play truant from school 
and was then sent to a private tutor, a woman. He went to her 
one day wearing his sister's high-heeled white slippers. Later 
attended public school but played truant again and was whipped 



CASE NOTES: MISCELLANEOUS 349 

by the principal. Next attended for one year a remote rural 
school of six pupils taught by the mother. There was no serious 
problem that year. At 12 he attended school in a large city. 
The children called him "sissy" and "old woman." The princi- 
pal tried to get him to fight, but could not. At this time X-13 
"behaved like a hunted creature." The truant officer finally 
decided there was somethirtg wrong with the boy and permitted 
his mother to keep him at home. On a camping trip (age 12) he 
was with some boys who planned to go swimming: "Mother, I 
just could not undress before them, and they did not like it." 
Once more attended an isolated rural school taught by his 
mother. The other two boys in the school will play with him 
only when the mother supervises them and they can't avoid it. 
Recurrence of truancy. 

The mother is desperate and feels entirely helpless. She 
recently took the boy to a psychiatrist who is said to be giving 
him a hormone extract. The mother thinks effort should be made 
to "persuade the Rockefeller Foundation to establish a home 
for the treatment of born inverts." 

One cannot say that this boy is not a "born invert." Cer- 
tainly invert tendencies appeared very early and have persisted 
in their typical course in spite of all efforts of the mother, school 
principal, truant officer, and psychiatrist to stem the tide. But 
there are several environmental factors that cannot be dismissed: 
the subject is the youngest child in the family and was born when 
his mother was approaching the menopause; the father is said 
to be an extreme introvert and not affectionate with his children; 
the mother is too affectionate and very emotional; much of the 
subject's life has been spent in isolated places where there were 
no boys except his older brother. 

X-14 

X-14 is a male high-school teacher, very old-womanish and 
censorious. 

Exercise 1234567 Total 

M-F score: -4 +1 -2 -32 -42 -1 -3 -83 



350 SEX AND PERSONALITY 

Only one third of college women test as feminine as this subject. 
His score is 3 S.D. below the mean of male teachers. 

X-14 was one of five children, the youngest except for one 
sister. He had two years in college before he began to teach. He 
is married. In Exercise 4 he checks "very much" for 14 of 17 
anger items and for 12 of 18 disgust items. In the ethical 
attitude section he rates every act in the list as either "extremely 
wicked" or "decidedly bad." Among those which come in for 
the most extreme censure are swearing, gambling, and talking 
back to the teacher! In Exercise 5 he expresses dislike for all 
the occupations listed except those of interior decorator, jeweler, 
missionary, and teacher. He claims much interest in reli- 
gion, social life, and domestic arts, little or none in mechanics, 
science, music, art, politics, or pets. His special hobby is 
philanthropy. 

X-15 

This is a 26-year-old druggist and chemist who stands in strik- 
ing contrast to X-14. 

Exercise 1234567 Total 

M-F score: +16 -3 +6 +60 +116 -1 -1 +193 

The score is more than 2.5 S.D. above the mean of males between 
20 and 30. The scores on Exercises 4 and 5 are extraordinarily 
masculine. 

X-1S was an only child. He is a college graduate and unmar- 
ried. His hobbies are swimming and shooting. He professes 
much interest in travel, sports, and science, average interest in 
mechanics, social life, literature, politics, domestic arts, and 
pets, and little or none in religion, music, and art. 

X-16 

Although engineers are our most masculine-scoring group, there 
are exceptions to the rule. The most feminine score we have 
encountered in this occupation was earned by X-16. 

Exercise 1 2 3 4 S 6 7 Total 

M-F score: -6 +1 +1 +35 -40 -12 -1 -22 



CASE NOTES: MISCELLANEOUS 351 

This score is nearly 100 points less masculine than the mean of 
graduate engineers and is near the mean of college women 
athletes and women doctors. However, the score on Exercise 4 
is more masculine than that of the average man. 

X-16 was an only child. He began to earn his own living at 
the age of IS and put himself through school. He was tested at 
29, after returning to the university for graduate work in engineer- 
ing. He had previously worked as a carpenter, boiler maker, and 
lawyer. He professes much interest in travel, sports, religion, 
mechanics, literature, music, science, politics, and pets an 
unusual combination of masculine and feminine tendencies. 

It is especially interesting that a follow-up of this case five 
years after he took the test disclosed that he never completed 
his engineering course. He attended intermittently for three 
years and left without his degree to work as a civil engineer. He 
is now married. 

X-17 

X-17 is a radio engineer, aged 28, who has the most masculine 
score of the engineers we have tested. 

Exercise 1234567 Total 

M-F score: +2 +1 4-56 +126 +7 -1 +191 

This score is near the 100th percentile of college males for 
masculinity. The score on Exercise 5 is extraordinarily mascu- 
line, and that of Exercise 4 decidedly so. 

X-17 rates himself as having much interest in mechanics and 
science, more than average interest in outdoor sports, and little 
or no interest in religion, literature, music, art, and politics. We 
have known this subject since his early youth. He has always 
been socially diffident and interested only in the most masculine 
of pursuits. He is married. 

X-18 and X-19 

X-18 and X-19 are engineers whose M-F scores are almost 
identical, but who differ greatly in interests. 

Exercise 1234567 Total 

X-18 +6 +1 +4 +54 +68 + 2 -3 +132 

X-19 -3 +10 +59 +76 -11 -1 +130 



352 SEX AND PERSONALITY 

Although there is a markedly negative correlation between 
M-F score and interest in religion (approximately .40), X-18 
engages in Sunday School work as his chief hobby. He rates 
himself as equally interested in religion and science. He has 
little or no interest in music, art, or pets. The score on Exercise 7 
suggests a somewhat introverted personality. 

X-19 represents an extreme case of the one-track mind which is 
extremely inventive and gifted in things mechanical. He was an 
only child whose hobby through childhood was photography. He 
was 25 years old when he took the test. He owns and runs a 
small plant in which he designs and manufactures special kinds 
of apparatus and appliances. He is extremely nonsocial. At one 
period he suffered a nervous breakdown and spent several 
months in a mental hospital. 

The only one of the 12 subjects in which he professes much 
interest is science. He has little or no interest in sports, social 
life, politics, or domestic arts, and less than average interest in the 
other activities. He illustrates the predominantly masculine 
trait of having one great interest in life to the exclusion of all 
others. 

X-20 and X-21 

X-20 and X-21 are two feminine-scoring ministers. 

Exercise 1234567 Total 

X-20 -5 -3 +5 -20 -62 -2 -3 -90 

X-21 -7 +11 -35 -18 +5 -3 -47 

X-20 is 47 years old. He was the oldest of six children, five 
boys and one girl. He has been married 20 years and has three 
children. He has much interest in travel, religion, and social 
life, little or none in sports, mechanics, or pets. His score is more 
feminine than the mean of adult women and about 2 S.D. below 
the mean of Protestant ministers. There is no evidence that his 
feminine attitudes have made his marital adjustments difficult. 

X-21 is also a Protestant minister, aged 65. Before entering 
the ministry he was in turn a clerk, factory worker, mechanic, and 
assistant foreman. He was the oldest of four children and has 



CASE NOTES: MISCELLANEOUS 353 

always been rather frail. He has been married 28 years and has 
one daughter. His hobby is men's choral work. The score is 
nearly one S.D. more feminine than the mean of Protestant 
ministers and approximates that of Who's who women. 

X-22 and X-23 

X-22 and X-23 are very feminine-scoring stenographers, aged 
22 and 26. 

Exercise 1234567 Total 

X-22 -18 -3 -30 -60 -64-9 -4 -188 

X-23 - 5 -2 -36 -41 -48 -24 -1 -157 

X-22 had two younger brothers and no sister. Her father died 
when she was still a child, and the children were reared by their 
mother. The subject had one year in college before going to 
work. She has much interest in travel, religion, music, and 
domestic arts, little or none in mechanics, art, or science. Her 
score is about 2.5 S.D. more feminine than the mean of female 
officer workers. 

X-23 was the only girl in a family of five children, her brothers 
all being older than herself. Her highly feminine attitudes may 
be in part the result of what we have called the "contrast 
effect"; that is, the tendency of parents and brothers to exert a 
strongly feminizing influence upon the family's one girl, especially 
when she is the youngest child. 

X-24 

X-24 seems to be an extreme example of the unemotional, 
apparently "hard-boiled," type of masculine woman. She is 26 
years old and married. 

Exercise 1234567 Total 

M-F score: -2 -5 +60 +34 -4 -1 +82 

This score is about 4 S.D. above the mean of adult women. The 
responses in Exercise 4 were especially masculine. Only 4 
of the 17 anger items are checked "very much," and not one 
of the 20 fear items. Only tobacco chewing in the disgust list is 
checked for "very much," and no other for "much." Not one 



354 SEX AND PERSONALITY 

of the pity items is checked for "very much" and only four for 
"much." The subject does not consider any of the acts in the 
censure list as "extremely wicked." 

X-24 has one sib, a younger brother. She finished high school, 
then attended business college and worked as a stenographer 
until she was married at 20. Her hobby is playing baseball. 

X-25 and X-26 

These subjects are both women in their thirties whose M-F 
scores are 218 points apart. 

Exercise 1234567 Total 

X-25 -9 -1 -3 +24 +14 - 5 -4+16 

X-26 -28 -3 -29 -35 -92 -14 -2 -203 

X-2S, the more masculine subject, was the oldest of three girls 
who were reared by their father. She graduated from high 
school, married when she was 22, and has three children. Her 
hobby is playing baseball. She has much interest in sports, 
social life, science, and politics, average interest in travel, 
mechanics, literature, music, art, domestic arts, and pets, and 
none in religion. 

X-26 scores about 3 S.D. below the mean of her decade group. 
She was the second of seven children, six girls and a boy, the 
latter younger than herself. She graduated from high school 
and worked as a stenographer until she was married at the age 
of 19. She has since taken a course in dressmaking. Her hobby 
is choir singing. She professes much interest in religion and 
music, none in mechanics, and average in the other nine fields. 

X-27 

This case is an example of the ultrafeminine, old-fashioned 
spinster in her fifties. 

Exercise 1234567 Total 

M-F score: -7 -2 -22 -69 -26 -21 -4 -151 

X-27 was the eldest of seven children. She attended business 
college after graduating from high school, but has never worked 
outside the home. She describes herself as a "home body." 



CASE NOTES: MISCELLANEOUS 355 

She has much interest in religion, little or none in sports, science, 
mechanics, politics, art, or pets. In exercise 4 she checked "very 
much" for 14 of the 17 anger items, for 17 of the 20 fear items, 
and for 13 of the 15 pity items. In the censure test she checked 
19 of the 28 listed acts as "extremely wicked." 

X-28 and X-29 

These are female octogenarians, differing in M-F score by 
about 2.5 S.D. Both are widows. 

Exercise 1234567 Total 

X-28 -29 -12 +2 -84 -11 -1 -135 

X-29 -19 -9+7 -22-5 +2-46 

X-28, the more feminine, was the sixth of thirteen children. 
Her father was a high-school teacher who taught his children at 
home. Without other formal education the subject taught 
school, became a social worker, and edited a small paper. She 
is the mother of seven children. Her hobby is religious work 
and she is also much interested in literature. 

X-29, the more masculine, was one of seven children, including 
three brothers older and three sisters younger than herself. She 
attended college two years, taught school, and later managed a 
farm. She is the mother of four children and now widowed 
after 50 years of married life. Her hobby is raising horses and 
dogs. She professes much interest in travel, literature, music, 
science, politics, and pets, and average interest in all the other 
fields listed. In breadth of interests she presents a striking 
contrast to X-28, a fact which is doubtless connected with the 
difference in M-F score. 

X-30 

X-30 is an unmarried man, 27 years old. He was the youngest 
of three children and rates his health as below average. 

Exercise 1234567 Total 

M-F score: -9 -2 +7 +15 -22 -7 -2 -20 

Although the M-F score of this subject is near the mean of the 
group of passive male homosexuals, his " in vert " score (184) is 



356 SEX AND PERSONALITY 

well within the normal range. Since completing high school he 
has worked as a delivery boy, filing clerk, and bookkeeper. His 
favorite pastimes are dancing and movies. He has much interest 
in social life, music, art, and pets, little or none in mechanics, 
science, politics, domestic arts, or religion. 

X-31 and X-32 

These two subjects, whose scores are separated by 134 points, 
illustrate the masculine and feminine extremes for elderly males. 

Exercise 1234567 Total 

X-31 -16 +1 + 1 +15 -56 - 7 0-62 

X-32 -10 -1 +11 +59 +30 -14 -3 +72 

X-31 is a manufacturer, aged 64. He was the youngest of four 
boys and went to school only two years. He was married at 30 
and has one daughter. His hobbies are music, geology, philoso- 
phy, and collecting stamps and wild flowers. He has much 
interest in travel, music, science, and politics, little or none in 
sports, social life, or domestic arts. 

X-32 is 71 years old. He was one of six children, five boys and 
a girl, who were reared by their mother. He attended a country 
school, and later became a carpenter. He was married at 26 
and has ten children. His hobby is horseback riding. He has 
much interest in travel, sports, and mechanics, little or none in 
music, art, science, or domestic arts. 

X-33 

X-33 was an upperclassman in a university when he came to the 
senior author for counsel regarding his homosexuality. Age 23. 
Intellectually gifted and the author of several published poems of 
genuine merit. He was given Form A of the M-F test, the 
Bernreuter personality inventory, and the Strong test of voca- 
tional interests. 

Exercise 1234567 Total 

Form A -18 -1 -11 +62 -60 -7 +1 -34 

Application of the "I" key to Form A yielded an "invert" 
score of +689, which is near the median found for 77 passive male 
homosexuals. 



CASE NOTES: MISCELLANEOUS 357 

The Bernreuter scores were as follows: "neurotic tendency" 
and " introversion" both at about the 90th percentile for college 
men; " self -sufficiency," 39th percentile; "dominance," 17th 
percentile. 

On the vocational-interest test "A" ratings were secured for 
the occupations of artist, journalist, and advertiser; "B+" for 
lawyer, and "B" for doctor, psychologist, architect, real-estate 
salesman, and mathematician. All other ratings were "C." 

The subject reports "much" interest in literature, music, art, 
travel, and pets; "little or none" in outdoor sports, religion, 
mechanics, science, politics, and domestic arts. He has at 
various times worked as a clerk, counter-man, actor, tutor, and 
chauffeur. 

X-33 came chiefly for advice as to whether he should leave 
college. He was in a wrought-up and nervous condition which 
he attributed to the fear of detection and to the emotional strain 
from living in a dormitory with attractive males to whom he 
dared make no advances. He has a rather slight build and an 
attractive personality. Voice not unusual, highly sophisticated, 
rather cynical, no ability at all in sports. As a child was plump, 
liked to keep his hair long, and tried to attract attention to his 
looks. 

He has two brothers, aged 34 and 28, and a sister aged 36. He 
is therefore the youngest of the children by five years. The 
father was 34 and the mother 29 when he was born. The father 
is a contractor, described as stern, distant, and cold. As a child 
X-33 feared him and still harbors bitterness toward him; says 
he is the Babbitt type, coarse, untruthful, and despicable. He 
deserted the mother for another woman when the subject was 9, 
but the son's aversion to him goes back at least to the age of 5. 

The subject's description of his mother disclosed nothing out 
of the ordinary as to her personality. She is said to have dis- 
played only an average amount of affection toward her son; there 
was no unusual fondling; subject not bathed or dressed by mother 
beyond the age of 6. The mother had been told she could have 
no more children and on finding she was pregnant wanted the 
child to be a girl. 



358 SEX AND PERSONALITY 

X-33 says that his brothers are more talented than he and more 
favored by the parents. Much friction has existed between the 
subject and his brothers. 

X-33 was treated by his parents rather like a girl, was con- 
sidered a sissy by children, preferred to be with girls, played with 
dolls till about 5 years old, and had no boys' playthings. His 
sister was sweet and gentle and let him play with her beads and 
clothes. He dressed up often in her clothes till age 12 or 13, and 
at times till IS or 16. 

He is rather social and prefers friends who are older and have 
cultured, broad interests. He never joined the Boy Scouts or 
any kind of lodge. His ideal character is Leonardo da Vinci. 
His first sex information was from books at age 12. As an 
adolescent he was bored by women but interested in men. Was 
taught to masturbate at 12 or 13 and continued the practice, 
probably four times weekly, for years. His first homosexual 
experience occurred at the age of IS. A man accosted him in 
the library, took him out riding, and persuaded him to play the 
passive role in fellatio. The experience was enjoyed by the 
subject and was frequently repeated. His first experience in 
pederasty was at the age of 20. He prefers fellatio or intercrural 
connection. 

Recently X-33 forced himself to carry through intercourse with 
a girl. He had tried before but could not get an erection. With 
difficulty he succeeded this time and had an orgasm, but was left 
with a feeling of disgust. At present he practices fellatio " quite 
frequently." He has no feeling of guilt and does not want to be 
made "normal." On the other hand, he looks upon hetero- 
sexuality as unaesthetic and unclean. He wants to leave 
college and return to the stage "where I can find companionship 
with others of my kind without risking disgrace." 

Significant factors: youngest child; stern, cold father; indulgent 
older sister; desertion of mother by the father; lack of stimulus 
to play with children; transvestism not discouraged; early 
seduction. The case is typical of a large number of male inverts. 

We come next to 12 cases of "gifted children" (6 girls and 
6 boys) whose M-F scores deviate excessively from their sex 



CASE NOTES: MISCELLANEOUS 359 

norms. All of them are members of the California group of 
1,000 superior children who have been followed by the senior 
author since 1922. They will be designated as G-l, G-2, etc. 

G-l 

This is a 13 -year-old girl whose masculine score is near the 
100th percentile for her age. Her IQ at 7 years was 160. 

Exercise 1234567 Total 

M-F score: -22 -1 +8 +18 +62 -2 +2 +65 

G-l is the older of two daughters of superior parents. The 
father is a structural engineer. Engineering and mechanical 
pursuits have been the most common occupations among the 
male relatives on both sides of the family. 

G-l is outstanding because of her mechanical ingenuity and her 
all-round masculine interests. Her skill with tools was first noted 
when she was only 3J^ years old. At the age of 4^ she con- 
structed an airplane that was placed on exhibit in Berkeley, 
California. By the age of 7 she was making sailboats, wagons, 
kites, "scooters," and bows and arrows. At this age she could 
understand fairly complicated machinery and was able to 
construct devices from printed plans or diagrams. At 13 her 
interest in tools had not diminished and she was considering 
either engineering or aviation as a career. 

At 7 G-l was making collections of sea shells and birds' nests. 
By 13 she had collected 900 varieties of stamps. Throughout 
childhood she preferred boys as playmates and read boys' books 
almost exclusively. At 13, for example, she was reading adven- 
ture stories, West Point series, Tom Swift series, Trade Wind, 
We, Boy Scout books, Boys 9 Life Magazine, etc. At this time 
her mother writes: "Her playmates look for her to start some- 
thing, and the play she starts usually has to do with the army 
or navy. She wants to see every movie about West Point or 
Annapolis and in these days of 'peace talk' her interest in war is 
somewhat distressing, especially in a girl." 

All her school work is excellent, but she prefers science, 
physical education, and orchestra to other subjects. She plays 



360 SEX AND PERSONALITY 

in the school band, is on several athletic teams, and is an honor 
student. The interest she has developed in music is rather 
exceptional in one testing so masculine. 

Physically, G-l is rather small for her age. She is somewhat 
square featured and her legs show the signs of early rickets. Her 
movements are noticeably short and jerky and her voice is 
somewhat squawky and gutteral like that of an adolescent boy. 

G-2 

This is a 17-year-old college girl whose IQ at the age of 11 was 
140. 

Exercise 1234567 Total 

M-F score: -25 +1 +3 +28 +60 +5 -3 +69 

G-2 is an only child, the daughter of middle-aged parents who 
have thought more of the pleasure she could give them than of 
what they could do for her. Her father is a teacher. Although 
chiefly of French, German, and English descent, the subject is a 
direct descendant in the sixth generation of the famous Indian 
chieftain, King Philip. 

She has always disliked housework and is very fond of outdoor 
games, sports, swimming, and horseback riding. She is a 
member of several athletic teams. She has shown some mechani- 
cal interests but has not worked much with tools. By the age of 
13 she had developed considerable interest in music, and on 
entering college selected art as her major subject. However, she 
has no great interest in any of her studies, preferring social 
excitement, dancing, and dates with boys. She shows the effects 
of her home spoiling. Her dass marks, which were very superior 
throughout high school, have become mediocre. 

G-2 is feminine looking and pretty, the kind who would hardly 
be expected to make a masculine score. She reached puberty 
at the age of 12%. 

There are two facts of special interest regarding G-2. (1) Her 
father is extremely attached to her and insists on "dating" 
her to movies and other affairs much as if he were a suitor. (2) 
Her mother confesses her hatred of men in general; says she 
"never knew one who was really decent." 



CASE NOTES: MISCELLANEOUS 361 

A large part of the masculine score in Exercise 4 is due to her 
responses to the censure items; only one of the acts listed (exces- 
sive drinking) is rated as " extremely wicked" and only six as 
"decidedly bad." In Exercise 5 she expresses liking only for the 
occupations of architect, artist, building contractor, detective, 
draftsman, and forest ranger. 

G-3 

G-3 is a girl of IS whose IQ at 9 years was 144. 

Exercise 1234567 Total 

M-F score: -1 -2 -8 +20 +40 +8 +2 +59 

She is the youngest of three children, two girls and a boy. Her 
father, of Spanish descent, is a truck driver and was formerly a 
rancher. The mother taught school before her marriage. 

G-3 is exceptionally attractive and is much sought after by her 
girl friends. She is something of a leader, although she prefers to 
devote most of her spare time to reading and drawing. She reads 
chiefly adventure books, detective stories, essays, and books of 
travel. She is planning to become a commercial artist. 

She rather enjoys working with tools and dislikes dances, 
parties, and housework. She likes active sports moderately 
well, but is not an enthusiastic athlete. She sews well enough 
to make her own clothes, but does it only from necessity. Aside 
from reading and drawing, she is rather indifferent to all activities, 
masculine as well as feminine. Although she scores almost as 
masculine as G-l and G-2, her behavior seems to be appreciably 
less masculine. 

The next three cases are gifted girls who test extremely feminine. 
The reader will note that in a number of respects they stand in 
marked contrast to the three cases just described. 

G-4 

Exercise 1234567 Total 

M-F score: -32 -2 -4 -55 -56 -8 -5 -162 

G-4 was 14 years old when given the M-F test. Her IQ at 
9 years was 147. Her Stanford Achievement Test score at the 



362 SEX AND PERSONALITY 

latter age was exceptional in the fact that it showed a mastery 
of reading equal to the ninth-grade level, three full grades above 
her score in any other school subject. 

She is the eighth of nine children, six of whom are girls. Her 
father is a revivalist and missionary of Scotch and English 
descent. The mother's family seems to be of good stock but 
to have had little education. When G-4 was born her father 
was S3 and her mother 41 and it is possible that her oldish 
parents may have had a feminizing effect upon her. 

G-4 is greatly interested in English and dramatics and plans to 
become a teacher of English. Most of her leisure is spent 
writing poetry or short stories and in reading. She has made 
collections of poems and short stories. Her reading preferences 
are almost entirely feminine. She has shown little interest in 
active games and none in working with tools. She is deeply 
religious. Her only very low marks in high school were in 
business methods, practical arts, bookkeeping, and stenography. 
Her health has always been fairly good, but she is described by 
her teacher as somewhat nervous, fidgety, and self-conscious. 

It is interesting that this ultrafeminine girl is a direct descend- 
ant of David Crockett. 

G-5 

Exercise 1234567 Total 

M-F score: -12 -3 -18 -14 -102 -7 -156 

G-5 is 17 years old. Her IQ at the age of 11 was 147, at which 
time she made a markedly feminine score on the Plays, Games, 
and Amusements Test. At 17 she is a university freshman, 
majoring in French, but planning to be a kindergarten teacher. 
Her scores on the Iowa High School Content Examination are 
at the 95th percentile except in the science and mathematics 
section, where she stood at the 75 th percentile. 

Her father is a merchant and manufacturer, her mother a 
teacher. Among her ancestors are many teachers, ministers, and 
inventors. 

G-5 is described by her teachers as modest, unassuming, 
sensitive, sympathetic, and lovable. She early learned to 



CASE NOTES: MISCELLANEOUS 363 

sew, knit, and crochet, and has never shown any interest in 
sports or the use of tools. A record of her reading over a two- 
months period when she was 11 years old contains only titles of 
distinctively feminine books. 

G-6 

Exercise 1234567 Total 

M-F score: -21 -1 -8 -6 -114 -4 -154 

G-6 made the above score at the age of IS. Her IQ at age 9 
was 152. Both parents are of Swedish descent. The father runs 
a small lunch-box factory and does church work as a hobby. 
The mother, who also works in the factory, is interested in music 
and social work. The relatives are mostly farmers of the 
religious, conservative type. 

G-6 is the oldest of three children, all girls. She is a striking 
blonde, good looking and of attractive personality, though 
slightly overweight. She displayed musical talent at 4, began 
music lessons at 7, and has learned to play the piano well. She 
prefers to spend her leisure time reading, sewing, playing the 
piano, and doing housework. Typical of her reading at IS are 
Pride and Prejudice, Jane Eyre, Ramona, Atlantic Monthly, and 
Pictorial Review. She is much interested in dancing and dra- 
matics, and has had leading parts in several school plays. Since 
she was in the middle grammar grades she has been the recipient 
of many honors and offices. She swims well and plays volleyball 
and basketball, but these have never ranked among her strongest 
interests. 

At 9 her Stanford Achievement Test score rated her especially 
high in reading ability and knowledge of literature, and rather 
low in history and civics. At IS she stood at the 85th percentile 
of high-school seniors in total score, but only at the 65th percentile 
in mathematics and at the 75th in science. 

The characterizations of G-6 by her teachers read like the 
traditional descriptions of the old-fashioned, feminine tempera- 
ment: "One of the dearest and most dependable children I have 
ever known" "Always loyal and eager to help" "When a 
task is given her we know it will be done" "No paint, no slang, 



366 SEX AND PERSONALITY 

and mathematics are his best subjects in high school, languages 
his poorest. However, he scores in the best 5 per cent of high 
school seniors in all parts of the Iowa High School Content 
Examination. He would like to become an army engineer. 

Perhaps this boy's most outstanding trait is his solitariness and 
incapacity for social adjustment, which may be in large part the 
result of his speech defect. His difficulties in social adjustment 
have probably in turn accentuated his interests in mechanics. 
Things, he can understand; personalities can be so much more 
baffling, unpredictable, and disagreeable. 

G-9 

Exercise 1234567 Total 

M-F score: -6 +1 + 11 +58 +94 +6 +164 

G-9 is 19 years old, a postgraduate in high school. His 
Stanford-Binet IQ at 13 was 130, but as the test lacks "top" at 
this level his true IQ was probably near 140. At 18 he scored 
between the 97th and 99th percentile of high-school seniors on 
all four parts of the Iowa High School Content Examination. 

G-9 is the second of three children, the oldest a boy, the young- 
est a girl. His parents, both deceased, were divorced when he 
was young and he lived with his mother until her death, when he 
was 13. The father was a civil engineer who "invented" during 
his spare time. The mother was a teacher before marriage. 
Males on the paternal side of the family have been chiefly 
engineers and architects, on the maternal side chiefly lawyers and 
businessmen. The son exhibited unusual manual skill and 
mechanical ingenuity at an early age. His ability to mend 
broken apparatus and to invent new contrivances was always 
far beyond his years. 

This subject, like the preceding one, is strongly inclined to 
solitude but without marked maladjustment. He is shy, seldom 
enters into school activities, and has made few friends. He likes 
to study atlases and his reading includes encyclopedias, adventure 
stories, and the dictionary, which he is said to have read dear 
through. His high-school work has been mostly indifferent. He 
repeated two history courses, chemistry, and advanced algebra, 



CASE NOTES: MISCELLANEOUS 367 

but made A's in solid geometry, mechanical drawing, and one 
course in English. He expects to be an aeronautical engineer. 

The next three subjects are at the opposite extreme in M-F 
rating from the three just described. Their scores are in the 
neighborhood of 2 S.D. below the high-school norms and not 
far from the mean of the passive male homosexual group. On 
the "invert" scale, however, they all score within the normal 
range. 

G-10 

Exercise 1 2 3 4 S 6 7 Total 

M-F score: -16 +1+5-12-14+8 0-28 

"I" score: -87 -34 +89 -153 -54 -60 -136 -435 

G-10 is a 15-year-old boy whose IQ at age 9 was 143 . He is of 
Russian- Jewish descent. His father is a businessman whose 
hobby is dramatics. The mother was a teacher and is interested 
in painting. His one sibling is a sister three years his junior. 

G-10 has always had frail health and much of the time it has 
been necessary to limit his school attendance to a half-day 
schedule and to keep him in bed the rest of the day. Even with 
this handicap he was a high-school senior when tested. His 
acceleration of three years dates back to the primary grade, 
which means that it has been impossible for him to compete in 
plays and games on equal terms with his classmates, especially 
as he is small for his age. His natural interest in books has 
doubtless been intensified by these circumstances. Puberty 
was late, if we may judge by the fact that his voice did not change 
until he was 16 years old. 

When G-10 was 9 and in the sixth grade his Stanford Achieve- 
ment Test score equaled the norm for the twelfth grade in literary 
information, the eleventh grade in history, civics, and general 
information, but only the eighth grade in science. At 15 he 
scored on the Iowa High School Content Examination at the 
95th percentile of high-school seniors in English, mathematics, 
and history, but only at the 75th percentile in science. He has 
always been rated low in mechanical ingenuity. His teachers 



368 SEX AND PERSONALITY 

describe him as exceptionally industrious, faithful, and ambitious. 
Socially he is well adjusted; was president of his dass in the 
eighth grade, and in high school was captain of the debating 
team and publicity editor of the high-school paper. Five years 
after taking the M-F test he won highest honors in the graduate 
school of law of a leading university. 

G-ll 

Exercise 1234567 Total 

M-F score: -12 -1+7 +35 -46 +2 - 2 - 17 

"I" score: -70 -19 +138 -21 +29 -75 -83 -101 

G-ll is a 17-year-old boy whose IQ at age 10 was 160. He is 
the oldest of three children. The father is a teacher and both 
parents are college trained. They live on a farm. 

G-ll made his highest achievement test scores in history, 
literature, and general information; his lowest in arithmetic, 
spelling, and mathematics. In no subject, however, was he 
below the highest quartile. Special talent in music was evidenced 
at the age of 4; he plays the piano well and is a member of the 
high-school orchestra. Both his mother and his teachers rated 
him very low in mechanical skill. He is characterized by 
teachers and field psychologists in such terms as "tractable," 
" quiet," "unassuming," "retiring," and "inclined to procrasti- 
nate." Although mathematics has not been his strongest 
subject, he plans to become a statistician. He devotes much 
of his leisure to reading philosophical works and composing 
music. At age 6 his masculinity score on the Plays, Games, and 
Amusements Test was below average. Like G-10, he was about 
two years retarded in reaching puberty. 



5 6 7 Total 

+ 2 -7 -4-12 
-108 -90 -91 -482 

G-12 was tested at the age of 16. His IQ at age 10 was 148. 
The "I" score indicates normality. He has a brother and a 
sister, both younger. The father is a teacher. Both parents are 



G-12 


Exercise 


1 


2 


3 


4 


M-F score: 


-16 








+ 13 


"I" score: 


-72 


-17 


+31 


-135 



CASE NOTES: MISCELLANEOUS 369 

college graduates and are interested in church work. Several 
ancestors on both sides have been ministers or church workers. 

G-12 has shown more than average interest in music since 
early childhood, but has not had musical instruction. For years 
he preferred fairy stories to any other kind of reading, but now 
enjoys books of travel and adventure. He is well adjusted 
socially and has been elected to numerous class offices. He takes 
pride in being popular. When he was just a boy his parents 
often left him in charge of the younger children and taught him 
to cook. He is said to be unusually dependable. He became 
interested in girls at the age of 14. Is noticeably small for his 
age and very fastidious about his clothes. 

G-12 intends to be a physician and scored A for that vocation 
on the Strong Vocational Interest Test. At the age of 16 his 
scores on the Iowa High School Content Examination were in 
English and history near the average of high-school seniors, 
but at the 85th percentile in science and mathematics. His 
interest in the last-named subjects appears to be greater than is 
common with boys who score so feminine on the M-F test. 

Although the fragmentary case notes presented in this 
chapter do not afford a safe basis for generalizations, they 
illustrate concretely a number of relationships statistically 
established in earlier chapters and suggest others which call for 
further investigation. For example, it is no accident that 
masculine-testing subjects of either sex are much more likely 
than subjects at the feminine extreme to be interested in sports, 
science, and mechanics, or that the more feminine subjects 
are more likely to be interested in religion, music, art, and 
domestic arts. It appears, indeed, that these positive and 
negative interests are also likely to characterize the subject's 
near relatives. The highly masculine female seems especially 
likely to have a father who is an engineer or has otherwise 
displayed an interest in mechanics. The daughters, perhaps 
also the sons, of ministers or other religious workers seem more 
often than others to score highly feminine. A few cases suggest 
that frail health or low energy level may operate as an eff eminiz- 
ing influence in males. Other cases among those reviewed 



370 SEX AND PERSONALITY 

emphasize the need of investigation along a number of lines: 
the influence of early and late puberty on M-F status, the 
personality differences between active and passive female 
homosexuals, the M-F status of persons given to unrestrained 
sexuality (rapists and nymphomaniacs), the vocational success 
of feminine-scoring engineers and of masculine-scoring ministers 
or artists, et cetera. 



CHAPTER XVI 
SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 1 

The greater part of our treatment of the M-F data so far has 
dealt with inter- and intra-sex differences in total score of the 
M-F test or in scores of the separate exercises. In other words, 
our approach has been primarily quantitative and statistical. 
Men and women have been compared with respect to mean score 
on the test as a whole, and subgroups within a given sex have 
been compared with each other in the same manner. It is 
interesting and important to know that mean M-F scores of the 
men and women of our adult population differ by about 122 
score points, or two and a half times the S.D. of either distribu- 
tion; that these distributions overlap to the extent of about 
10 per cent; that something like ten out of a thousand of each 
sex reach or exceed the mean of the other sex; that the M-F score 
is related in complex ways to age, education, intelligence, occu- 
pational classification, expressed interests, and the domestic 
milieu; that the scores have yielded such and such correlations 
with scores on a variety of other personality tests; or that 
extreme scores by individual subjects are likely to connote 
exceptional types of social and sexual adjustment. The value 
of an approach which is productive of such generalizations is 
so obvious as to need no defense. The quantitative method has 
justified itself. 

But the psychologically minded reader will want to know more 
about sex temperaments than quantitative methods of the 
totalitarian kind just referred to can tell him. He is unwilling to 

1 The material reported in this chapter was assembled by Mr. Horace G. Wyatt 
under the direction of the senior author and is based on the item-by-item responses 
of the 550 subjects used in the standardization of the two forms of the M-F test. 
The chapter as it appears was prepared by the latter from a 200-page summary 
written by Mr. Wyatt. The authors are greatly indebted to Mr. Wyatt for the 
psychological acumen with which he performed his task. 

371 



372 SEX AND PERSONALITY 

stop with sex differences expressed in terms of score-points; he 
would like to know what is behind these points that can be 
expressed in a psychological rather than a statistical vocabulary. 
We may think of this reduction of data to psychological cate- 
gories as constituting the qualitative as contrasted with the 
quantitative approach. We regard the two methods as present- 
ing a genuine contrast, notwithstanding the fact that qualitative 
methods also at times utilize quantitative concepts. In the 
present case, certainly, our task would be very incomplete if we 
confined our treatment of M-F differences to just those facts 
that can be expressed in terms of score points. We must 
reexamine our data with a view to reducing them if possible to 
categories that are more strictly psychological than the majority 
of those which we have thus far employed. 

The material to be presented in this chapter seeks to elicit 
from the responses made by several hundred males and females 
in the various exercises of the M-F test any important sex 
differences in interest, sentiment, disposition, and temperament 
that a comparison of the responses of the sexes to the individual 
items may reveal. Of the seven exercises in the M-F test, the 
last four inquire directly into the emotions or sentiments awak- 
ened, or professedly awakened, by certain actual or assumed 
situations or objects. The first three exercises proceed indirectly, 
on the assumption that sex divergences in associative tendencies 
(Exercises 1 and 2) and in actual knowledge (Exercise 3) are 
likely to derive from differences in sentiment and interest. 

The exercises were administered to subject groups of both 
sexes, representative of grade, high-school, and college students, 
and nonacademic adults, the first two groups predominating. 
As the populations tested were not quite the same throughout, 
they are specified with each exercise as it is considered. In the 
case of each exercise item tabulations were made separately 
for the male and female groups and the item-by-item differences 
between the sexes have been used as a basis for the delineation of 
the outstanding features of the sex temperaments. 

No daim, it should be premised, is here made that the exercises 
or much less the items taken separately are suitable for individual 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 373 

diagnosis, i.e., for determining with confidence from an indi- 
vidual's response the "maleness" or the "femaleness" of that 
subject's mentality. What we are concerned with are general 
tendencies that actually distinguish the sexes in the large, and 
may thus be claimed to constitute "femininity" or "mascu- 
linity" without any assumption with regard to the relative 
influence of genetic and environmental factors as causal agents. 

EXERCISE 1: QUADRUPLE-CHOICE WORD ASSOCIATION 

As the reader will recall, the subject's task in Exercise 1 is to 
underline that one of the four given associates following a 
stimulus word which seems "to go best with" the latter. Each 
form of the M-F test contains 60 stimulus words, making 120 in 
all. 

In this and all the other exercises with the exception of Exer- 
cise 3, the present study of sex temperaments is based upon the 
weighted scores of the individual items which are given in 
Appendix IV. The weighted scores were used rather than unit 
scores in order that magnitude as well as direction of the sex 
differences could be taken into account. 

The population groups used in the derivation of weights for 
Exercise 1 included roughly 125 pupils of each sex in grade seven, 
100 of each sex in the junior year of high school, and SO college 
students of each sex; in all about 550 subjects. These are the 
subjects who furnish the data for our sex comparisons on Exer- 
cise 1. 

For purposes of convenience, and also because they appeared 
more promising of results, the responses to Exercises 3, 4, and 5 
were examined before those to Exercises 1 and 2. When Exercise 
1 was taken up a number of hypotheses had already been reached 
with regard to the nature of some of the more important psycho- 
logical differences between the sexes. These hypotheses, based 
upon an examination of the responses of male and female groups 
to Exercises 3 to 5 were utilized as a basis for forecasting the sex 
preferences of associations in Exercise 1. It will be interesting to 
see how adequately the explanatory hypotheses, empirically 



374 SEX AND PERSONALITY 

arrived at from the responses in certain exercises, explain the sex 
differences in exercises of a very different type. 

Predictive Analysis. Evidence from the analysis of Exercises 3 
to 5 justified the forecast that if word preferences correspond to 
object preferences, 1 males would be found choosing more often 
than females word associates symbolizing aggression, activity, 
adventure, extradomestic (scientific, physical, political, business, 
and economic) affairs, and customary male occupations; while 
females would be found preferring words for domestic things or 
happenings, articles of dress or personal adornment, words for 
colors, and for artistic technicalities or occupations; words indic- 
ative of active, especially maternal, sympathy, and of customary 
female occupations. 

An attempt was therefore made to forecast for Exercise 1 which 
of the quadruple-choice responses would be chosen more often by 
males and which by females. Comparison later of these forecasts 
with the tabulated data disclosed the fact that correct predictions 
were more than twice the number chance alone would have 
yielded, and that a majority of the responses wrongly classified 
had either a zero weight or a very small weight in the opposite 
direction to the forecast. But as even the most carefully selected 
groups are never completely representative, and the examiner 
cannot be completely familiar with their circumstances, and as, in 
any case, each response is conditioned not only by its relation 
to the stimulus word, but also by the counterattractions of three 
other response words, it would be strange if he did not frequently 
fail in the difficult task of calculating the competing factors for the 
four responses. That twice the chance number of predictions of 
sex preferences proved correct, and that it was chiefly where sex 
preferences were less decided that prediction failed, is so much 
support for the hypotheses as to dominant sex-distinguishing 
traits upon which the predictions were based. 

The above digression from our main task (the examination of 
empirical data) may or may not have been worth the time it 
cost. It does show that the psychologist, with definitely formu- 

1 This assumption may not always hold; e.g., some words are more affected by 
one sex than by the other. 



SEX TEMPERAMENTS AS REVEALED B7 THE M-P TEST 375 

lated hypotheses as to how the sexes differ, can predict with con- 
siderable success how men and women will differ in their responses 
to particular test items. It should be pointed out, however, that 
the same thing can be done with no small degree of success by 
young college students who are not psychologists and whose ideas 
about sex differences have never been clearly formulated. We 
have already seen that it is possible for relatively unsophisticated 
subjects to fake their scores so as to make them typical of members 
of the opposite sex. Truly the psychologist has no monopoly of 
psychological insight into the characteristic differences which 
distinguish men and women. Let us turn, however, to an exam- 
ination of the actual data. 

Evidence from the Most Sex-contrasting Choices. If and in so 
far as our assumed sex differences hold and govern choices in 
verbal associations, it is where the choices of four presented asso- 
ciates show most sex divergence that we might fairly expect the 
assumed differences in sex traits to help determine them; and 
where, of the four words, two appeal to contrasting sex pro- 
pensities, we might expect to find between them, other things 
equal, the widest divergences in the weighted scores. Conversely, 
where the weight spans are widest, we should anticipate fewest, if 
any, records contradicting our expectations as to sex preferences. 

Confining our scrutiny, then, to those cases, rather less than a 
third of the whole, in which to the same stimulus word the 
difference between the sex responses is of six or more units (e.g., 
stimulus word "Pole" with response "north, 4," "telephone, 
+3," a difference of seven units) we may ask which distinctive 
associations are referable to our assumed sex distinctions; and 
whether any results contradict or significantly supplement what 
we might expect. 

Out of the total of 120 items the 40 which exhibit these pro- 
nounced contrasts are given in Table 64. In interpreting this 
table it has to be remembered that a sufficiently pronounced 
preference of one sex for one response may automatically throw 
another response on to the other sex, since fewer scores of the 
former sex remain to be shared among the other responses. Our 
interpretation, therefore, is adequate if one of the responses in a 



376 



SEX AND PERSONALITY 



contrasting pair is " explained." The table shows the contrasting 
responses with the suggested category of reference, or explanatory 
trait, assigned in every case to which our hypothesis seems most 
dearly to apply. The "explained" 1 response is underlined. A 
query in the column for "explanatory characteristic" indicates 
that the explanation is doubtful. 

TABLE 64. INTERPRETATION OP CONTRASTED ASSOCIATIONS 



No. 


Stimulus 
word 


Male 
response 1 
(plus) 


Female 
response 
(minus) 


Explanatory characteristic 


Form A 
1 
7 


POLE 
CASE 


telephone 3 
bottles 4 


north 4 
doctor 3 


Scientific 
Domestic and sympathetic 


8 
10 
13 

14 


POST 
JACK 
BRACE 

FLY 


fence 4 
money 3 
bitS 

airplane 3 


gate 2; letter 2 
toy 4 
support 4 

bird 3 


Outdoor v. domestic 
Domestic v. business 
Mechanical (tools) 

Scientific. Adventure 


18 

20 
21 
22 

23 
24 
29 


BOOK 

MOON 
FLESH 
DANGER 

MODEST 
FRESH 
GARDEN 


paper 2 
print 2 

light 4 
meat 3* 
caution 3 

bashful 5 
meat 3* 
weeds 3; 


read 4 

night 4 
color 3 
accident 3 

shy 5 
cool 4 
flower 5 


Literature for the female side; 
mechanical ("things it is made 
of or with") for the male 

Art and decoration 
? 

? 
? m 
Decorations for the female, and 


30 
33 

46 
27 
52 


EMBRACE 
BABY 

FACE 
SPOON 
MARRIAGE 


vegetables 3* 

lover 3 
cry3 

pretty 3 
pet a; soup 2* 
divorce 4 


arms 3 
darling 3 

powder 4 
silver 4 
children 2 


outdoor occupation for the 
male side 
? 
Maternal sympathy; the male 
records what strikes him most 
about a baby at a distance 1 
Personal embellishment 
Decoration 
Maternal 


59 


VAIN 


impossible 3 


proud 4 


The female takes the more inti- 


Form B 
2 
3 

4 


LIVE 
MUFFLER 
NEEDLE 


wire 3 
T3 
compass 4 


breathe 4 
warm 4 
sew 2; eye 2 


mate and personal sense of the 
word 

Mechanical (tools) 
Mechanical v. domestic 
Scientific for the male, and 


5 

10 


NOTE 
SHIFT 


mark 3 
football 4 


letter 3 
gear 3 


domestic for the female 
Domestic and personal v. busi- 
ness 
Male addiction to outdoor ac- 


12 
15 


BUM 
LIPS 


hobo 3 
kiss 3 


tramp 3 
mouth 5 


tivities and emulation throws 
even the mechanical "gear" 
on to the female side 

? 



1 Weights for female responses are minus, for male responses plus. 

1 That males tend to prefer "food" words is evidenced sporadically throughout the seven 

1 The word "explained" is here used in the loose, popular sense. A type of 
response is of course not really explained by merely referring it to a behavior 
category. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 377 



TABLE 64. INTERPRETATION OF CONTRASTED ASSOCIATIONS. (Continued) 



No. 


Stimulus 
word 


Male 
response 1 
(plus) 


Female 
response 
(minus) 


Explanatory characteristic 


Form B 
18 
19 


TOMBOY 
SNAKE 


girlS 
reptile 3 


boy 3 ; rough 6 
grass 3 


Scientific term 


23 
25 


CHILD 
POWDER 


gun 3; explo- 
sion 5 


girl 5; cun- 
ning 4 
rouge 2; face 7 


Decoration for the female, and 
adventure for the male 


27 
34 


SWEETHEART 
WAIST 


kiss 4 
belly 4 


friend 2; 
lover 2 
belt 2 


Unaesthetic and so disliked by 










females 


37 


NASTY 


filthy 3 


horrid 3 


? 


48 
49 

51 
52 


ELECTRIC 
ART 

DARKNESS 
MAKE 


generator 4 


light 3 
artist 3 

afraid 4 
dress 4 


Scientific 
Art for the female, and me- 
chanical contrivances for the 
male 

Domesticity v. business 


gallery 3 


kiss 2 
money 3 


54 

55 

56 


SLEN 
CHAR 

WORM 


skinny 3 
snake 4 


graceful 4 
grace 3 

squirm 3 


Aesthetic v. unaesthetic 
The more aesthetic v. the more 
"exciting" word 
Outdoor pursuit 


fish 3 



1 Weights for female responses are minus, for male responses plus. 

That males tend to prefer "food" words is evidenced sporadically throughout the seven 
exercises. 

Out of 40 most marked contrasts between the sexes, 26, it 
will be noted, seem readily explicable in the terms of our hypothe- 
sis. Of the 12 queried items, excess of caution prevents us from 
explaining some on a similar basis. There is no instance of a sex 
preference reversing expectations. Of the records, then, of these 
sex-contrasting cases, most support and none contradicts our main 
hypothesis, leaving as one might expect a minority of cases in 
which the distinctive sex motives are not immediately obvious, 
and in which no fresh sex difference emerges. 

Sex Characters of the Same Associates in Different Contexts. 
We must of course evade the snare of taking the response abso- 
lutely, and considering it "unreliable" if a given word happens to 
obtain a male score in one context and a female score in another. 
In choosing one of four given "associates" the subject has in mind 
not only its relation to the stimulus word, but its relevance to the 
stimulus in comparison with that of three rival associates each 
associate has to pass not only a qualifying but a competitive 
examination, with different competitors in each case. But 
where, in spite of variety of context, a word adheres to one sex, we 



378 SEX AND PERSONALITY 

may suspect some rooted sex preference for that word or for what 
it symbolizes, or some rooted aversion to it in the opposite sex. 

The number of associate words presented more than once in the 
present exercise was 26, and of these 2 were presented five times, 
2 four times, 8 three times, and 14 twice. In 7 out of these 26 
associates, the "preferring" sex differs with the occasion; but on 
the whole a sex tends to maintain its preference through varying 
contexts. Among the 12 offered more than twice, 5 are persist- 
ently or predominantly male (money, meat, love, soldier, kiss) and 
4 (picture, mother, baby, happiness) persistently or predominantly 
female. Five of 12 items have the same sex trends here as 
obtained with the dass they belong to in the M-F test as a whole 
(e.g., words of the dass economic are generally male, etc.) . These 
5, with sex preference and response dassification, are as follows: 
money (male, economic), meat (male, foods), soldier (male, 
activity and adventure), picture (female, aesthetic), and mother 
(female, same sex preference). The remaining 4 indude love and 
kiss (both male) ; and baby and happiness (both female) . Clearly 
the female score for baby is explicable occupationally. We for- 
bear a posteriori explanation of the others. The point to note is 
that most cases of reiterated distinctive sex preference correspond 
to sex trends elsewhere evidenced and accord with our main 
hypothesis. None contradicts it. 

Analysis According to Classes of Associate Words. In this 
case we shall be taking the associates absolutely, that is, without 
considering along with them their stimulus or associate context; 
for though the sex score for any particular associate is obviously 
influenced in each case by that context, yet we may fairly assume 
that over a sufficient population particular extraneous influences 
cancel out, exposing for the two sex groups their sex distinctions. 

In Table 65 the quadruple-choice assodates have been dassified 
into 17 groups suggested by a separate study of free associations 
for which space is not found in the present volume. 1 The scores 
are the totals of the weighted scores for words of the dass con- 
cerned; e.g., for the words for "parts of the body" selected by a 

1 WYATT, H. G., Free word association and sex difference, Amcr. J. Psychol., 
1932, 44, 454-472. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 379 



higher proportion of males than of females the weighted male pref- 
erence scores totaled 4 and 12 in Forms A and B respectively, and 
for those with the female excess the total female preference scores 
totaled 1 and 18 respectively. 

TABLE 65. TOTAL WEIGHTS OF SEVENTEEN RESPONSE CLASSES 



Response class 


Male 


Female 


Total 


Form 
A 


Form 
B 


Form 
A 


Form 
B 


M 


F 


1. Body parts 


+ 4 

+ 2 
+ 5 
+ 3 
+26 
+21 
+ 17 
+22 
+ 7 
+ 5 
+ 5 
+ 14 
+19 
+ 11 
+ 2 
+ 12 


+12 


+ 3 
+ 2 
+ 2 
+30 
+25 
+ 19 


+ 5 
+13 
+ 9 
+ 12 
+ 9 
+ 6 


- 1 
- 8 
- 4 

-19 
-13 
-10 
-10 
-25 
- 7 

- 4 
- 2 
- 1 
- 7 
- 3 
- 9 


-18 
-12 
-10 
-17 

-18 
- 9 
- 6 
-20 
- 2 
- 5 
- 4 

- 2 

r 

- 8 
-10 
-19 


16 

2 
8 
5 
28 
51 
42 
41 
7 
5 
10 
27 
28 
23 
11 
18 


19 
20 
14 
17 
19 
31 
19 
16 
45 l 
9 
5 
8 
4 
6 
15 1 
13 1 
28 1 


2. Dress 


3 . Personal adornment 


4. Aesthetic 


5. Colors 


6. Domestic 


7. Outdoor 


8. Activity and adventure 


9. Emotional 


10. Religious 


11. Social 


12. Business, economic, and political. . . 
13. Scientific and mechanical 


14. Foods 


IS. Personal male 
16. Personal female . 
17. Personal neutral 



1 If we exclude emotional words with a markedly sexual significance, the scores for emo- 
tional words run *Hi (instead of Jfg) and KB (instead of 8 %o). giving total M 22, P 39; 
and for personal male words M 13, P 7; personal female, M 1, F 7; personal neutral, M 14, 
F26. 

Table 65 shows a female distinctive associative trend toward, or 
tendency to think of, words for articles of dress and personal 
adornment and colors, and words of aesthetic appraisal; a male 
tendency toward words for outdoor phenomena, of activity and 
adventure, of science and machinery, and for foods. Under the 
categories "parts of the body/' "domestic," "emotional," 
" religious," "social," "business-economic-political" and "per- 
sonal " in general, sex differences appear slight, perhaps negligible. 

Thus this analysis of the associates taken absolutely, that is, 
without considering their stimulus and associate context, agrees 



380 SEX AND PERSONALITY 

with our previous findings as to the female addiction to aesthetic 
objects, personal embellishment, and dress; and the male to 
adventure and to extradomestic (outdoor, business, economic, 
and political) and scientific and mechanical affairs. There is 
again a rather marked male preference for foods, and there is a 
marked female preference for colors; in addition, there are certain 
neutral areas. The "domestic" associates reveal no marked sex 
distinction. A reference to details showed, however, that all the 
" food " words in the exercise happen to fall in Form A, and secure 
a considerable male preference; while the Form B domestic 
associates (comprising no "food" words) and the "maternal" 
words in Form A (infant, baby, toy, mother) score predominantly 
female. In domestic matters exclusive of foods females appear 
the more interested. 

Summary. Whichever method of analysis has been used, 
whether (1) prediction based on hypothesis (derived from an 
analysis of Exercise 3) as to characteristic sex interests, (2) the 
most sex-contrasting responses, (3) sex constancy in associations, 
or (4) sex conformity in preferences with those evidenced in 
separately administered free association exercises, we find 
females picking more often than males upon terms for domestic 
things or happenings, for kindly and sympathetic activities, and 
for articles or tokens of adornment; males more often than females 
upon words relating to physical science, machinery, outdoor 
pursuits, and terms suggestive of excitement and adventure, and 
rather less predominantly, upon political, business, and com- 
mercial words. 

EXERCISE 2: INK-BLOT ASSOCIATION 

In each form of the M-F test 18 ink blots, or drawings, are pre- 
sented and the subject is asked to underline one of the four words 
following each (" the word that tells what the drawing makes you 
think of most"). 

The populations used in the analysis of responses to Exercise 2 
were the same as for Exercise 1 . We have tabulated the responses 
of the subjects in two ways, one the less illuminative, taking the 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 381 

associations absolutely; the other by considering the more sharply 
contrasted associations in relation to one another and to the 
stimulus object. The first tabulation, which is self-explanatory, 
serves the former purpose: 

Form A 

Male preferred associations: 
+6: coil. 

+ 3: pipe; tadpole; hat; ax; man hanged; cannon; kettle. 
+2: brush; target; flagpole; bell; fish; nose; head. 
+1: tombstone; pear; snow shoe; centipede; boat; moon; torch; 

fence; lovers; baby; saxophone; bat. 
Female preferred associations: 

4: mail box; ball bat; coal bucket. 

3: inkwell; chopper; dish; chimney. 
2: comb; sword; smoke; thread; incense. 

1: spoon; teeth; ship; cloud; dancers; scarecrow; tassel; Indian 

club; jug; idol; snake; flower; star; bowl; cup. 
Neutral: 

jar; ham; cradle; cup; fish; mirror; ring; tire; bird house; letter "E"; 

tree; Indian; babies; bottle; whip; goat; face; a man. 
FormB 

Male preferred associations: 
+ 5: jack; butterfly. 
+ 3: glasses; skate; buoy. 
+2: columns; bomb; chain; cow; boat; flower; hook; staff; dagger; 

stickpin; baby; valve. 
+ 1: footprints; diamond; vase; cone; lady; tree; hook; boat; hat; 

stump; sword. 

Female preferred associations: 
6: bow. 
-5: lady. 
4: basket. 

3: lady; question mark; sled. 

2: vase; bow; funnel; exclamation point; door; flower; stove. 

1: candle; gate; jar; tie; light; rock; ship; horn; couch; deer; horse; 

coffin; flame; snake; worm; pyramid; shadow. 
Neutral: 

fireplace; flask; pan; snake; trees; ribbon; bush; mushroom; haystack; 
umbrella; club; stick; string; airplane. 

Throwing the associated words into classes we may compare 
the sex preference scores as follows: 



382 SEX AND PERSONALITY 

Forms A and B combined: 

1. Terms associated with mechanical and scientific interests. 

Male associations: coil; saxophone; valve; jack. 

Total male, 4 

2. Terms connected with outdoor activities or adventure. 

Male associations: butterfly; glasses (field); skate; buoy; bomb; 
cow; boat; dagger; tree; boat (again); sword; stump; cannon; 
man hanged; flagpole; boat (third time); tadpole; torch; tomb- 
stone; snow shoe; centipede; fence; bat; pear; target; fish; ax. 

Total male, 27 

Female associations: sled; gate; rock; ship; deer; horse; snake; 
pyramid; worm; ball bat; sword; star; scarecrow; snake (again). 

Total female, 14 

3. Terms connected with domestic occupations. 

Male associations: pipe; kettle; bell; baby; baby (again); hat. 

Total male, 6 

Female associations: coal bucket; chopper (food); dish; comb; 
thread; spoon; jug; bowl; cup; basket; funnel; door; stove; 
candle; tie; couch; jar. 

Total female, 17 

4. Terms connected with aesthetic experience or personal adornment. 

Male associations: flower; diamond; vase; stickpin. 

Total male, 4 

Female associations: bow; vase; bow (again); flower; incense; 
flower (again); dancers; tassel; comb. 

Total female, 9 

5. Unclassified terms. 

Male associations: column; chain; hook; staff; footprints; cone; 
lady; brush; lovers; nose; head; moon. 

Total male, 12 

Female associations: mail box; inkwell; lady; lady (again); ques- 
tion mark; exclamation point; light; horn; coffin; flame; shadow; 
chimney; smoke; teeth; cloud; Indian club; idol. 

Total female, 17 

The results of the above treatment accord with previous 
generalizations as to the relative interests of the sexes. Further 
comment is unnecessary. 

We present next, in Table 66, the most strongly contrasted 
associations. 

Seemingly in every case, except perhaps for Item 17 of both 
forms, the distinctive occupations of the sexes determine the con- 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 383 
TABLE 66. CONTRASTED SEX ASSOCIATIONS IN INK-BLOT TEST 1 



Item 


Male 


Female 


Comment 


Form A 
1 


Pipe* +3 


Mailbox -4 


The male smokes a pipe 


2 


Tadpole +3 


Ball bat -4 


Outdoor interest v. a specific female 


3 


Hat +3 


Inkwell -3 


game 
Sedentary clerical occupation 


6 


Ax +3 


Chopper 3 


Kitchen v. woodpile 


12 


Coil +6 


Chimney 3; 


Scientific or mechanical v. domestic 






Smoke -2; 








Thread -2 




17 

Form B 
5 

7 


Kettle +3; 
Nose +2 

Jack +5 
Boat +2 


Coal bucket -4 

Funnel -2; 
Horn -1 
Basket -4 


? 

The male changes tires 
Domestic v. adventure 


10 


Butterfly +5; 
Glasses 


Bow 6 


Female adornment v. outdoor pur- 
suit 


17 
18 


(field) +3 
Skate +3 
Buoy -f3; 


Sled -3 
Lady -5 


? 
Scientific and outdoor v. sex-to-sex 




Baby +2; 

Valve +2 




tendency 











With sex divergence span of 6 up, e.g., the span for mailbox 4, pipe +3 is 7. 
1 The associate to which the comment applies is underlined. 

trasts. The comments suggest the determining factors in each 
case. In Item 18, Form B, "baby" is thrown on to the male side 
by the fact that the female, more discriminative in this case 
through specific experience, would never associate "baby" with 
the shape depicted. 

Summary. Females pick more often upon response words 
associated with domestic occupations, aesthetic experiences, and 
personal adornment; males more often upon words most obviously 
connected with machinery, physical science, and outdoor pursuits. 
The most sexvcontrasting responses evidence a preference for 
words connected with common occupations of the responding sex. 



384 SEX AND PERSONALITY 

EXERCISE 3: INFORMATION TEST 

Each form of this exercise consists of 70 sentences (i.e. 9 140 in 
all) to be completed by underlining that word out of four supplied 
that gives the correct information required. It is a quadruple- 
choice information-completion exercise, success in which calls 
primarily for knowledge of fact and would seem, therefore, to 
depend upon the experience and interests of the subject. 
Responses were analyzed for the same populations as in the case 
of the two previous exercises. 

We shall briefly summarize the results of our scrutiny before 
considering the detail. 

Within the range of the 140 items set and the subjects who 
responded to them, a comparison of the sex-preference scores 
indicates that: 

1. Females show more knowledge 

a. About domestic occupations. 

b. About matters of personal or household decoration, adornment, 

and etiquette. 

c. About literature (fiction), music (or rather musical technique), 

and conventional taste in colors. 

d. Where the topic is one that appeals to one's active sympathy 

or feelings of tenderness (including in particular the maternal 
emotions). 

2. Males show more knowledge 

a. Of extradomestic facts and events, political, social, economic, 

and business. 

b. Of physical and scientific facts. 

c. Of exploit, adventure, invention, whether in fact or fiction. 

d. Where the topic is one that appeals to the pugnacious, aggres- 

sive, or vigorously active tendencies. 

We can now take the evidence in detail. 

Analysis of the Most Sex-contrasting Responses. This 
analysis is limited to the correct responses carrying plus or minus 
weights of 4 or larger. 1 They are as follows: 

1 As will be seen in Appendix IV, M-F weights were assigned to all the responses 
in this test, the incorrect as well as the correct, and also to omissions. Our 
analysis, however, takes account of the correct responses only. The assumption 
behind our present analysis is that on the average a subject tends to answer more 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 385 

Form A 
Greater knowledge by males (weights +4 or higher) : 

6. The most gold is produced in Alaska New York Tennessee 

Texas 
21. The boomerang is an animal plant tool weapon 

24. The "Rough Riders" were led by Funston Pershing Roose- 

velt Sheridan 

25. The vessel which overcame the Merrimac was the Connecticut 

Monitor Old Ironsides 

31. The Erie Canal is in Canada Ohio New York Pennsylvania 
34. The mossy side of a tree is usually on the east north south 

west 

40. The proportion of the globe covered by water is about J Ji 

% 

43. Mica is an explosive food mineral vegetable 

47. Shinny is played with bats clubs nets racquets 

48. When water freezes it contracts expands does neither 

62. The altitude record for airplanes is about 10,000 ft. 20,000 ft. 

40,000ft. 60,000ft. 
66. Beam scales illustrate the principle of buoyancy elasticity 

leverage magnetism 

Greater knowledge by females (weights 4 or larger): 
1. Marigold is a kind of fabric flower grain stone 
5. Pongee is a kind of cloth drink flower game 

32. Red goes best with black lavender pink purple 

41. The turquoise is blue red white yellow 

44. Blue clashes worst with brown gray pink purple 
46. A buffet is used for books clothes dishes food 
51. The amethyst is green purple white yellow 

In the "male " list, Items 6, 21, 31, and 43 demand a knowledge 
of physical or scientific facts; 21, 24, and 62 are directly or closely 
related to adventurous activities; all the items except perhaps 47 
have to do with the world beyond the home; and 21, 25, and 62 
are topics of particular appeal to aggressive and enterprising 
natures. 

In the " female " list, five of the seven items carrying a weight as 
large as 4 for the correct response have to do with personal 

correctly on matters in which he is more interested or absorbed. Whether he 
also tends to prefer a response word signifying what he is most interested in, 
irrespective of its being "right" or "wrong," is a separate question not here 
considered. 



386 S&X AND PERSONALITY 

adornment. The exceptions call for information about the mari- 
gold and a buffet, both related to the domestic situation. 
The corresponding items in Form B are as follows: 

Greater knowledge by males (weights +4 or higher) : 

4. Robert E. Lee surrendered to Grant Lincoln Sheridan Sher- 
man 
13. Sun eclipse is caused by the shadow of the earth Jupiter Mars 

the moon 

16. True silk comes from goats plants sheep worms 
21. Paul Jones was a general sea-fighter statesman writer 
24. The telegraph was invented by Bell Edison Morse Whitney 
3 3 . The large prehistoric reptiles were called anthropoids crustaceans 
diatoms dinosaurs 

36. Of these woods the hardest is cedar oak pine walnut 

39. The blacksmith of the gods was called Jupiter Mercury Perseus 

Vulcan 

40. Tides are caused by winds rotation of the earth attraction of 

the moon ocean currents 

41. A shipping term is FOB RST SOS TNT 

42. "Pi "is equal to .6666 .7853 1.453 3.1416 

46. Heating a piece of iron makes it larger lighter smaller stronger 
55. The diameter of the earth in miles is about 2,000 5,000 8,000 
25,000 

58. Light travels per second about 8,000 miles 25,000 miles 186,000 

miles 240,000 miles 

59. Among the allies of Germany, in the World War, was Belgium 

Bulgaria Roumania Russia 

60. The "Block System" is used in carpentry mining railroading 

surveying 
66. A noted general in the Mexican War was Burnside Hooker 

Jackson Taylor 

Greater knowledge by females (weights 4 or larger): 
23. A mixture of red and blue pigments gives brown green purple 

yellow 
32. A baby usually begins to walk alone at about 6 mos. 12 mos. 

20 mos. 24 mos. 

37. Topsy was Arabian Eskimo Indian Negro White 

38. A chiffonier is used for books clothes dishes silverware 
50. Heliotrope is the name of a drug flower gem tree 

Examination of the Form B lists shows that the items divide 
according to the same principles which govern their division in 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 387 

Form A. Of the 17 in the male list, no fewer than 10 or 11 may 
be classed as scientific. Four are historical, and three of these 
have to do with war. 

Analysis of Less Divergent Responses. We shall now consider 
the items in which the sex predominances, though noticeable, 
were less conspicuous; namely, those with weights of 2 or 3. If 
actual differences in sex characteristics are revealed by the differ- 
ences in the character of the correct responses, then where the 
response differences are most pronounced the sex differences are 
most clearly revealed, and where the response differences are less 
pronounced other factors besides sex differences enter in. As the 
smaller score weights are statistically less reliable than the larger, 
the analysis of less divergent responses has been limited to Form A. 

Form A 

Greater knowledge by males (weights +1 to +3): 

3. The Yale is a kind of hammer lock screen wrench 
7. The earth moves around the sun in 7 days 30 days 180 days 
365 days 

11. Most of our anthracite coal comes from Alabama Colorado 

Ohio Pennsylvania 

12. The number of players on a baseball team is 7 9 11 13 

15. Peat is used for fuel pavement plaster road-making 

16. Marco Polo was a famous king philosopher traveler warrior 

17. Tokyo is a city of China India Japan Russia 

28. A shilling is worth about 25 cts. 50 cts. $1.00 $5.00 

36. Turpentine comes from coal petroleum trees whales 

37. The number of persons on a jury in the U.S. is 8 12 16 18 
39. The chief cause of the tides is the attraction of the moon 

planets sun stars 
42. A plant "breathes" chiefly through its bark leaves roots 

twigs 

49. The Roman numeral C equals 50 100 500 1,000 
53. The length of a brick is 6 in. 8 in. 10 in. 12 in. 

56. Barometers are used to measure air pressure heat humidity 

rainfall 

57. Lobo was the name of a bear crow fox wolf 

59. The number of ordinary steps in a mile is about 1,000 2,000 
5,000 10,000 

63. Limestone originated from granite marble sand shells 

64. An animal that suckles its young is the alligator shark snake 

whale 



388 SEX AND PERSONALITY 

65. A birthright was sold for a mess of pottage by Cain Esau 

Isaac Judas 
68. A decisive Revolutionary battle was Gettysburg New Orleans 

Valley Forge Yorktown 
Greater knowledge by females (weights 1 to 3) : 

2. Things cooked in grease are boiled broiled fried roasted 
4. We should drink tea from the cup saucer spoon 
8. A stately dance of colonial days was the minuet polka two- 
step waltz 
10. Beethoven is known as a composer painter poet singer 

13. Eggs are best for us when deviled fried hard-boiled soft- 

boiled 

14. A loom is used for cooking embroidering sewing weaving 

19. Daffodils are grown from bulbs cuttings seeds shoots 

20. The baby found in the bulrushes was Jacob Jesus Moses 

Paul 

22. Minnehaha means falling leaves laughing waters running 

brooks whispering pines 

23. A correct expression is I have dove I dived He dove 

26. About A.D. 1750 men's sleeves had bands lace-ruffles stiff 

cuffs stripes 

27. A food with much the same food substance as rice is beans peas 

meat potatoes 
30. "Speak for yourself" was said by Annabel Lee Evangeline 

Juliet Priscilla 
33. Baby gets its first tooth at about 6 mos. 12 mos. 15 mos. 

18 mos. 
38. The Madonna is a favorite subject for music paintings poetry 

stories 
45. A dinner hostess seats the guest of honor at her left opposite 

right 
50. Some think "moon over the right shoulder" means death rain 

sickness wish fulfillment 
52. Ruth and Naomi are known for their devotion hatred pity 

rivalry 
55. "Mennen's" is the name of cold cream perfume collar 

talcum 
58. "Charades" is a running game game of chance guessing 

game kissing game 

60. "Peter Pan" was written by Barrie Kipling Mark Twain 

Stevenson 

61. Babies should be weaned at about 3 mos. 6 mos. 12 mos. 

2 yrs. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 389 

67. A character in "David Copperfield" is Betty Uriah Heep 
Sindbad Oliver Twist 

70. A famous portrait painter was Rosa Bonheur Mozart Rey- 
nolds Rubens 

In the male list, Items 3, 7, 11, IS, 36, 39, 42, S3, 56, 59, 63, and 
64 demand a knowledge of physical or scientific facts; the others 
relate to war (68), adventure (16 and perhaps 57), geography 
(17 and 28), civics (37), and baseball (12). 

In the female list, Items 2, 13, 14, 27, 33, and 61 are closely 
connected with domestic occupations or home life; 4, 8, 26, 45, 
and 55 have to do with personal adornment or etiquette. Of the 
remaining items, 20, 30, 52, 60, and 67 are from literature, mostly 
from works of fiction in prose or poetry. Items 10, 19, 38, and 70 
have to do with art, music, or flowers, 20 and 52 call for biblical 
information, and 50 is a romantic superstition. 

Summary. The information test, scored for correct response, 
gives valuable clues to the diverging interests and preoccupations 
of the sexes. The masculine key quality brought out by the test 
is the aggressive, adventurous, enterprising, outwardly directed 
disposition; the tendency to pugnacity (in the wider sense) and 
self-assertiveness. Related to this quality is the greater interest 
in things physical and scientific. The outstandingly feminine 
traits are the actively sympathetic, the inwardly directed dis- 
position; the maternal impulse and the tender feelings; concern 
with domestic affairs, personal adornment, art, and literature. 

For masculinity there appear to be two crucial questions: Does 
the individual under consideration show a taste for adventure, 
rivalry (up to the fighting point) and aggression? And secondly, 
does he evince an interest in things outside the domestic circle? 
Of these the former is the more significant of masculinity. After 
this in order of importance we may come to such questions as: 
Does the subject show an interest in physical and mechanical 
phenomena? and perhaps still lower in the scale: Does he show an 
interest in social, historical, and business affairs? 

For femininity on the other side we may venture a correspond- 
ing order: Does the individual concerned show a taste for, or an 
interest in, topics or events that arouse the sympathetic emotions 



390 SEX AND PERSONALITY 

(those that call for active sympathy) or the maternal interest in 
infant life? Does he (or she) show a pronounced interest in the 
intimacies of home and personal relationships rather than in the 
less personal relationships of external affairs? After this in 
order of importance may come such questions as : Does the subject 
show an interest in personal adornment? and perhaps still lower 
in the scale: Does he or she show an interest in art and literature 
(especially fiction)? 

As regards art and literature it should be noted that all our 
information exercise suggests is that the girl knows better than 
the boy who were the great artists, what common technical terms 
in music signify, and about certain color relationships and con- 
ventional judgments of color harmony. We cannot safely infer 
any superiority in artistic emotional appreciation. And as 
regards literature the evidence rather suggests a greater attention 
on the part of the girl to works of fiction and to poetry; one can 
draw no inference about her taste for literature in general. 

EXERCISE 4: EMOTIONAL AND ETHICAL RESPONSE 

Exercise 4 falls into six sections, of which four concern occasions 
of anger, fear, disgust, and pity respectively; the fifth considers 
occasions of moral reprobation; and the last compares preferences 
for alternative situations or types of occupation. The analysis 
of Exercise 4 was based upon data from the following populations : 
148 boys and 172 girls in the seventh grade, 116 boys and 128 girls 
in the junior year of high school, and 50 college students of each 
sex; in all, 664 subjects. 

Anger Reactions. Here the subject is presented with a number 
of assumed situations calculated to arouse the emotion in some 
degree, and with four degrees of emotional intensity: very much 
(VM ), much (Af), little (L), or none (N) ; and is invited to record 
the particular degree of that emotion which the assumed situation 
provokes, or would provoke, in him or her. 

The first method of analysis adopted was to pair VM with M 
as representing a tendency to considerable emotion, and L with N 
to indicate defect and to contrast the combined paired records. 
But it is dear that on the excess side VM and on the defect side 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 391 



N indicate more and less emotion respectively than M and L\ for 
instance, a "minus 2 VM " with a "minus 2 M " represents a 
greater tendency to keen emotion than a "minus VM " with a 
"minus 3 M ." No one of course can tell how much more emotion 
TABLE 67. SEX DIFFERENCES IN ANGER RESPONSE 



M-F test items 



Excess 
[2 VM + M ) 



Defect 



Span 



Form A, order for female excess over male defect: 

1. Seeing people disfigure library books 6 

2. Seeing boys make fun of old people 7 

3. Seeing someone laugh when a blind man runs into an 

obstacle -6 

4. Seeing someone cheat in an examination 4 

5. Being called lazy -4 

6. Hearing your political views ridiculed 1 

7. Being snubbed by an inferior 4 

8. Being blamed for something you have not done . . 2 

9. Seeing a person laugh at a cripple 3 

10. Hearing someone make fun of your clothes . . 2 

11. Being called stupid - 2 

12. Being unexpectedly slapped on the back as a joke . 2 

13. Being deceived by a supposed fnend 2 

14. Being called a thief 

15. Being disturbed when you want to work 

16. Seeing an honest official thrown out by politicians. . 

17. Seeing someone trying to discredit you with your 

employer +2 

Average span between female excess and male de- 
fect (85 - 5 + 17) 

Form B, order for female excess over male defect: 

1. Seeing a person treated unfairly because of his race 9 

2. Being socially slighted 5 

3. Seeing an innocent man punished for another's crime 6 

4. Hearing your friends unjustly abused ... . 7 

5. Seeing a man sitting in a car with old women standing 7 

6. Being marked unfairly in an examination. ... 6 

7. Hearing your religion ridiculed 5 

8. Being honked at by an automobile 4 

9. Being deceived by an enemy 2 

10. Hearing someone misspell or mispronounce your 

name 2 

11. Being slapped in a quarrel 2 

12. Seeing an animal cruelly beaten 2 

13. Being unreasonably prevented from doing what you 

want to do 1 

14. Being cheated in a business deal 

15. Being called homely 

16. Being called a liar 

17. Being called by a nickname you don't like 

Average span between female excess and male de- 
fect (100- 1 + 17) 



+6 



+4 
+ 5 
4-4 
+ 7 
+ 3 
+2 
+ 2 
+ 2 
+ 1 
+ 1 



-1 
-1 

-1 



+ 7 
+ 6 
+ 3 
+ 3 
+ 3 
+4 
+ 1 
+ 3 

+ 3 
+ 2 
+1 



-1 






12 
10 

10 
9 
8 
8 
7 
4 
S 
4 
3 
3 
2 


- 1 

- 1 

- 3 
4.7 



14 

12 

12 

10 

10 

9 

9 

5 

5 

5 

4 
3 

2 

- 1 




5.8 



392 SEX AND PERSONALITY 

is signified by a VM than by an M response; and the differences 
may vary between subjects and in the same subject on different 
occasions. But for practical purposes we may recognize the 
existence and the importance of this difference by giving VM and 
N double weight. Hence VM and M were taken together, and 
L and N together, the VM and N scores being doubled before 
adding them to the M and the L scores. The sum of the VM and 
M weights will be called the "excess" score, the sum of the L and 
N weights the "defect " score. Table 67 gives the items in order 
of decreasing "excess" score of female over male subjects. 

We may contrast the items in which the recorded differences are 
most pronounced with the items in which little difference appears 
and ask what is common to the former and absent, or present in 
less degree, in the latter. The items at the extremities of the 
scale represent the strongest contrasts. In Form A two of these 
items, the second and third, representing a high degree of female 
as compared with male annoyance, are instances of very unsym- 
pathetic behavior, while no such cases occur at the other end of the 
series. Item 9, the only other of this type in Form A, is at about 
the middle of the scale. 

In Form B four out of five instances of very unsympathetic 
treatment where sympathy is demanded come at the top of the 
scale, namely, items 1, 3, 4, and 5; 1 the exception being twelfth. 
Except item 12, all these cases in Form A represent human beings 
receiving some considerable injury when they need humane treat- 
ment. Thus six cases of singularly unsympathetic treatment of 
human beings when sympathy is meet are also cases of great 
difference between the sexes in the anger expressed, the female 
expressing the keener emotion. This corroborates the hypothesis 
of the greater tendency to active sympathy in the female, at least 
where human beings are the object. 

The other two items at the head of the scale, the first and 
fourth in Form A and the sixth in Form B, suggest that girls are 
more annoyed by school offenses than boys are ("disfiguring 
library books " and " cheating in an examination," "being unfairly 
marked in an examination") and proportionately less angered 

1 The numbers here refer to the order of items in Table 67. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 393 

than males by extraschool injustices, since two of the only three 
items in the whole series against which males record more anger 
than females include " being discredited with one's employer ' ' and 
"seeing an honest official thrown out of office unfairly." The 
former evokes the greatest degree of male anger recorded in 
the section. In Form B the only case near the top of the sex 
difference scale we have not accounted for is the female excess of 
anger over "being socially slighted" (item 2). The female sex 
is the more concerned about social, the male about business, 
relations. 

The responses in this section accordingly support the view 
that the female is the more actively sympathetic sex, and is more 
concerned in her nearer relationships than the male, being pro- 
portionately more affected by school and home and "social" 
vexations, and less by business or extradomestic troubles. 

A tendency is seen here, as in all the other sections of this 
exercise, for females to record in general a higher degree of emo- 
tion. Of the 34 items relating to anger, in 26 the excess score is 
on the female side and in only one on the male side. On the other 
hand, the defect scores are correspondingly male. 

Fear Reactions. This section of the test is like the preceding 
except that the subject's response indicates degree of fear aroused 
by the situations specified. 

The rank order of the fear items according to decreasing excess 
of female over male scores is shown in Table 68. 

The influence of cultural bias on responses would seem likely 
to be stronger for fear responses than for anger responses. For 
the female sex is allowed by society to be timid, and the male is 
expected to be brave. The young man fears to be afraid, and the 
habit of refusing fear expression is bound to weaken his responses 
in this section considerably. But there is no such sex convention 
for anger. Hence, we should expect the fear scores to yield us 
wider spans of sex divergence than the anger scores, irrespective 
of any really wider divergence in emotional intensities between the 
sexes. A glance at the table confirms this expectation. The 
widest "anger" spans are 12 and 14 for Forms A and B, while 
the widest "fear" spans are 23 and 22. The averages of female 



394 SEX AND PERSONALITY 

TABLE 68. SEX DIFFERENCES IK FEAR RESPONSE 



M-F test items 


Excess 
(2 VM + 3f ) 


Defect 

(L + 2N) 


Span 


Form A, order for female excess over male defect: 
1. Floods 


-13 


4-10 


23 


2. Garter snakes 


-12 


+ 11 


23 


3. Burglars 


-12 


+ 10 


22 


4. Bulls 


-12 


+ 9 


21 


5. Being lost 


-10 


+ 10 


20 




- 9 


+ 11 


20 


7. Graveyards at night 


- 8 


+ 6 


14 


8, Being in a closed room .... 


- 7 


+ 6 


13 


9. Deep water 


- 6 


+ 7 


13 


10. Windstorms 


- 4 


+ 6 


10 


11. Lightning 


- 4 


+ 5 


9 


12. End of the world 


- 3 


-|- S 


8 


13. Negroes 


- 5 


+ 2 


7 


14. Heart trouble 


- 3 


+ 3 


6 




- 1 


+ 3 


4 


16. Becoming deaf or blind . . . .... 


1 


+ 3 


4 


17. Pain 


- 2 


+ 1 


3 







+ 1 


1 


19. Insects 











20. Contagious diseases 


+ i 


2 


- 3 


Average span for Form A (221 3-4- 20) 






10.9 


Form B, order for female excess over male defect: 


-10 


+ 12 


22 


2. Fire 


10 


+ 11 


21 


3. Rattlesnakes 


-12 


+ 7 


19 




- 9 


+ 9 


18 




10 


+ 6 


16 




- 7 


+ 7 


14 


7, Guns 


- 6 


+ 7 


13 


8. Mice 


- 4 


+ 7 


11 


9. Toads 


- 3 


+ 6 


9 


10. Being on a ship . 


4 


+ 4 


8 




- 6 


+ 2 


8 


1 2. Crossing an open space 


_ 2 


+ 6 


8 


13. Ghosts 


4 


+ 3 


7 


14. Written examinations 


2 


+ 4 


6 


15. Knives 


- 1 


+ 4 


5 




2 


4- i 


3 


17. Horses 


- 1 


+ 2 


3 


18. Becoming crippled or totally disabled 


3 


_ i 


2 


19 Poisoning 


4. i 


o 


1 


20. Old age . 


+ 3 


3 


6 


Average span for Form B (193 7 * 20) 






9.3 



excess over male defect of expressed anger are 4.7 and 5.8 for 
Forms A and B, for fear, 10.9 and 9.3. 

We shall now examine Table 68 in detail. Our method of 
analysis was to take Form A first, and to scrutinize the first nine 
as against the last eight items with a view to eliciting suggestive 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 395 

f 

contrasts. Three contrasts suggest themselves. The items in 
the first group as compared with the second comprise perceptual 
events or objects that either (1) are more threatening to life and 
limb; or (2) are more sudden and occasional; or (3) demand a 
more immediate display of active courage; or have more than one 
of these attributes. This contrast is the more obvious the nearer 
we approach the extremes on the scale. "Floods," "garter 
snakes/' "burglars," and "bulls" in which the female excess of 
expressed emotion is greatest conspicuously fulfill all three 
conditions, and "punishment in the next world," "becoming deaf 
or blind," "insects," and "contagious diseases" items at the 
bottom of the scale do not. " Negroes," at the top of the lowest 
eight items, are not particularly threatening to life and limb, nor 
do they demand an immediate display of courage ; "heart trouble " 
is a state demanding careful treatment, not a quick display of 
courage; "automobiles" are hardly occasional; and "pain" is a 
felt state and not a perceived object at all. 

The reader may carry the comparison throughout the series. 
In general there is a descending trend according as the situations 
fulfill the conditions postulated. But since instinctive fear in man 
or higher animals is for objects or events that are perceived as 
threatening immediate danger to life and limb, especially events 
that are sudden and unfamiliar, and since such situations demand 
the immediate display of active courage, what the record suggests 
is not that there are objects or events that in themselves are more 
or less terrifying to one sex than to the other, but that things that 
are generally fearsome are, or are declared by them to be, more 
fearsome to females. 

A scrutiny of the Form B spans, while confirming, supplements 
this conclusion. For there are three out of the eight situations 
claimed as the most excessively fearsome by females, which do not 
fulfill our three conditions: namely, "being alone at night," 
"darkness," and "mice." A social reason for the female dread of 
darkness and solitude suggests itself. Girls are early taught, in 
some families more than others, not to go about alone, especially 
in the dark. Where this practice occurs the sex distinction is 
likely to be accentuated. "Mice" are also a special cultivated 
female terror. 



396 SEX AND PERSONALITY 

In summary, then, a scrutiny of the fear sections in detail 
suggests minor socially determined differences in the occasions of 
fear to the sexes; and a general tendency in the female sex to 
express fears more strongly, and that in proportion to the actual 
fearsomeness of the object. 

Disgust Reactions. The relevant data for our discussion of this 
section of Exercise 4 are given in Table 69. 

It will be seen from Table 69 that all items portraying actual 
male rather than female practices, and the two instances of 
behavior specifically mentioned as male, come high in the scale. 
These are Items 1, 2, 6; and 5 and 7. The female expresses much 
more disgust at them than the male does. That practices sanc- 
tioned among members of one sex and not the other cease to 
repel the former is a sufficient explanation. Though, since there 
are no specifically female disgusting habits included in either 
Form A or B this leaves us the questions whether and why males 
have the more disgusting habits. But our present exercise is not 
designed to answer these two questions. 

There remain in Form A two "extreme" items, namely, "soiled 
or ragged fingernails" and "untidy clothes" (Items 3 and 4) not 
thus explicable. But here again comes in a social factor. Amer- 
ican girls meditating matrimony also meditate their clothes and 
fingernails, on which they lavish far more care, cash, and cosmetics 
than is customary with males. The sex is therefore the more 
repelled by an offense against its idols. 

The score graduations on Form B are not disposed of so 
readily. We may select the first seven items for contrast with the 
last seven items on the scale. Of the five items that refer to 
clothes in this and the cognate section, four fall in the higher part 
of the scale; the keener female concern about dress would seem 
to explain greater disgust at dress disfigurements; this would cover 
Items 2 and 5. But the exception, the very low Item 13 in 
Form A, "food stains on clothing," defeats one. 

Items 2 in Form A and 1 in Form B indicate a female repug- 
nance against impolite language. 1 Cultural bias provides the 

1 The table on Ethical Discrimination Responses in a later section shows that 
"swearing" is more censured by females than by males. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 397 
TABLE 69. SEX DIFFERENCES IN DISGUST RESPONSE 



M-F test items 


Excess 
(2VM+M) 


Defect 

(L + 2N) 


Span 


Form A, order for female excess over male defect: 
1. A drunken man 


13 


4-11 


24 


2. Foul language 


12 


4-11 


23 


3, Softed or ragged fingernails - 


11 


4-12 


23 


4. Untidy clothes 


-12 


+ 9 


21 


5. Sagging socks on a man 


- 8 


-H2 


20 


6. Spitting in public 


-10 


+ 9 


19 


7. An unshaven man ... . 


- 9 


4- 


18 


8, Gum chewing ........ .,.,,, T 


- 7 


+ 11 


18 


9. Pimples 


- 6 


+ 8 


14 


10. Seeing a woman smoking 


- 8 


+ 5 


13 


11. Smell of decaying flesh 


- 7 


+ 6 


13 


12. Word "gent" used for gentleman. . . . 
13. Food stains on clothing 


- 6 
- 7 


+ 6 
4- 4 


12 
11 


14. Offensive breath 


- 6 


+ 5 


11 


15. Sight of slimy water 


- 6 


+ 4 


10 


16. Mushy food in your mouth 


- 4 


+ 5 


9 


17. Crooked teeth 


3 


+ 2 


5 


18. A butcher shop ... 





+ 2 


2 


Average span (266 -5- 18) . . . . 






14.8 


Form B, order for female excess over male 
defect: 
1. Words like "belly" or "guts" 


-18 


4-10 


28 


2. Soiled table linen . 


-12 


+13 


25 


3. A boy and girl petting 


-10 


+12 


22 


4. Tobacco chewing 


-11 


+ 11 


22 


5 Sight of dirty clothes 


-12 


+ 8 


20 


6. Sniffling 


- 9 


+ 10 


19 


7 Odor of perspiration 


- 9 


+ 8 


17 


8. Sight of pus 


- 7 


+ 5 


12 


9. Smell of onions 


- 4 


+ 8 


12 


10 Hearing a person belch 


- 7 


+ 4 


11 


11. Seeing a bull 


- 2 


+ 6 


8 


12 Sight of anyone vomiting . . 


- 4 


+ 3 


7 


13 Smell of cooked cabbage 


- 2 


+ 5 


7 


14. Wienerwursts or stuffed sausage 
15 Dirty neck or ears 


- 3 
- 9 


+ 4 
- 3 


7 
6 


16 Seeing a man smoking . .... ... 


- 3 


+ 2 


5 


17 A banana 


_ 2 





2 


18. Bad table manners 


- 7 


- 5 


2 


Average span (232-3-18) 






12.9 











398 SEX AND PERSONALITY 

reason: coarse language is accounted more offensive in the girl 
than in the boy. 

Item 4, " tobacco chewing/' aligns with Item 8, "gum chewing," 
as a disgusting practice, the former being more usual with males 
and also higher in the scale. 

There is also a tendency as in the fear section for the items 
which might be expected to provoke the keener emotion (which- 
ever the sex) to come higher in the scale. Conversely, a reason 
why the item "seeing a man smoking" is an apparent exception 
to the generalization that specifically male practices show the 
greater spans is that the practice is not repugnant enough to call 
forth much disgust in any case and is also socially sanctioned in 
men by women when it is not so sanctioned in women. 

As one would expect, it is the more intermediate items that 
most puzzle one. It is hard to say why "sniffling" (Item 6) 
should be comparatively more disgusting to females than, say, 
"dirty neck and ears" (Item 15). There remains Item 3, ''a boy 
and girl petting," high in the scale. If active physical sexual 
desire is stronger in males than in females, this may more effectively 
counterbalance moral repugnance, and cultural bias may also 
come in, for female offenses against "virtue" are the more 
condemned. 

In conclusion, it appears that there are many "nasty practices" 
which the male tends more than the female to condone or not to 
find disgusting; that the female sex is more particular about coarse 
or impolite language, about aesthetic accessories of person or 
dress, and about sexual morality. 

Pity Reactions. Our discussion of this section of the exercise is 
based upon the data in Table 70. 

A corollary of the tendency already mentioned for the female to 
express her feelings with the greater emphasis is a general tend- 
ency for sex-difference scores and actual emotiveness of a situation 
to correspond, the most alarming or disgusting situations showing 
the higher sex spans. This complicates interpretation, for the 
recorded sex difference in any item now appears as the outcome 
of two independent variables: (1) this general tendency toward 
female emphasis; and (2) any true sex differences specific to the 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 399 
TABLE 70. SEX DIFFERENCES IN PITY RESPONSE 



M-F test items 



Excess 
(2VM + M) 



Defect 

(L + 2N) 



Span 



Form A, order for female excess over male defect : 

1. An orphan girl 8 

2. An insane person 6 

3. A wounded deer 5 

4. An old person with a fatal disease 5 

5. An underfed child -5 

6. A bee that is drowning 4 

7. Very old people 4 

8. A baby bird whose mother is dead 3 

9. A dog that must be killed for biting people 4 

10. A man who is cowardly and can't help it 1 

11. Overworked children 5 

12. A fly caught on sticky fly paper 1 

13. A wounded soldier who must beg for a liv- 

ing -2 

14. Overworked horses 2 

15. A young person totally paralyzed 1 

Average span (113 + 15) 

Form B, order for female excess over male defect: 

1. A mouse caught by a hawk 9 

2. An old maid who has always wanted chil- 

dren 7 

3. A bird with a broken wing 8 

4. A fish caught on a hook 6 

5. An orphan boy 7 

6. A boy ambitious to go to college but cannot 4 

7. A wounded lion 5 

8. An underfed horse 6 

9. A toad being swallowed by a snake 3 

10. A dope fiend - 1 

11. A young person with a fatal disease. .. . 3 

12. A terribly ugly man 1 

13. A man who is physically a weakling 2 

14. A person with no will power 2 

Average span (139 -* 14) 



+ 7 
+ 6 

+ 5 
+ 4 
+ 4 
+ 4 
+ 4 
+ 5 
+ 3 
+ 5 
+ 1 
+ 5 

+ 2 

+ 1 

+ 1 



-HO 

+ 11 
+ 6 
+ 8 
+ 7 
+ 8 
+ 6 
+ 4 
+ 4 
+ 4 
+ 2 
+ 3 
+ 1 
+ 1 



15 
12 
10 

9 

9 

8 

8 

8 

7 

6 

6 

6 

4 
3 
2 
7.5 



19 

18 

14 

14 

14 

12 

11 

10 

7 

5 

5 

4 

3 

3 

10 



item. We have, in other words, always to remember that items 
in which the true specific sex differences may be the same may 
yet differ widely in their place on the scale of descending spans 
because of the first variable. 



400 SEX AND PERSONALITY 

The reinforcement of general emotiveness by a specific factor 
appears in the sex divergences in this section, when we compare 
the first five with the last six items in Form A, and again in Form 
B. On the ground of mere emotiveness the items "overworked 
children" (11 in Form A), "a young person totally paralyzed" 
(IS in Form A), and "a young person with a fatal disease" 
(12 in Form B) are surely just as distressing as "a mouse caught 
by a hawk," " a bird with a broken wing," " a wounded deer," or 
"a fish caught on a hook" 1 (items high in the scale). 

But a glance at the last three items adds the specific factor. 
They all instance some weak or delicate creature perceived 
at the moment as helpless and in visible distress. The first three 
items represent a distressing state, in which the helplessness and 
the distress and the weakness of the subject are far less immedi- 
ately perceptible. The female claims to be more moved by the 
sight of the weak and helpless in temporary but palpable distress. 

We note again certain exceptions: "an orphan girl" and "an 
old maid who has always wanted children," although representing 
distressing states rather than instances, appear as Items 1 and 2 on 
Forms A and B. But here the "normal" female excess over the 
male is accentuated by the additional specific factor of sympathy 
with their own sex. "An orphan boy" registers a slightly lower 
score than "an orphan girl" and falls considerably lower in its 
scale. 

Again, objects that are unsightly tend toward a low position 
("a dope fiend," "a toad being swallowed by a snake," "a 
terribly ugly man"), suggesting that pity is counteracted by 
repulsion or disgust. These explanations for what they are 
worth (and they at least offer corroborative evidence) cover all 
the most sex-differentiating items except 2 in Form A "an 
insane person." Why this should seem so much more pitiful to 
females than to males is not clear. Why should female pity 
exceed that of males here more than, say, in the case of " a young 
person with a fatal disease"? 

Summarizing, a comparison of the female and male scores 
suggests or indicates as distinctive sex characteristics: (1) a 

1 "A bee that is drowning 71 is less maternally appealing bees sting. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 401 



female sympathy (the maternal instinct?) for the weak and help- 
less visibly distressed; (2) a tendency for the female to express 
more pity for female distress; (3) a tendency to more female than 
male sympathy with things attractive in appearance as con- 
trasted with the ugly. 

TABLE 71. SEX DIFFERENCES IN DEGREES OF CENSURE 



M-F test items 


Excess 


Defect 


Span 


3 
Ex- 
tremely 
wicked 




Not 
really 
bad 


Form A* order of female excess over male 
defect: 
1. Boys fighting 


-8 


+ 7 


15 


3 


4-3 


2. Swiping fruit out of orchards 


-6 


+ 8 


14 


_3 


4-3 


3. Shooting rabbits just for fun 


-4 


-f-8 


12 


-2 


4-3 


4. Boys smoking before they are 21 . . 
5. Indulging in "petting". ... .... 
6. Going to bed without saying your 


-4 
-5 

-5 


+7 
+6 

+5 


11 
11 

10 


-1 

-2 

_2 


4-3 
4-3 

4-2 


7 Stealing a ride on a truck 


6 


+ 3 


9 


_2 


4-1 


8 Moderate drinking 


_4 


4.5 


9 


' 2 


4-2 


9. Not brushing your teeth 
10. Having fits of temper 
11. Boys teasing girls . 
12. Putting pins on the teacher's chair. 
13. Telling a he to avoid punishment. 


-4 

-2 
-3 
-4 
-3 
-3 


4-5 
4-7 
+ 5 
4-3 
4-3 
4-3 


9 
9 
8 
7 
6 
6 


-2 

-1 
-1 
-2 
1 


4-2 
4-3 
4-2 
4-1 
4-1 
4-1 


15. Picking flowers in a public park . 
16. Laziness 
17. Boy running away from home. . 
18. Breaking windows. . . 
19. Insulting the defenseless. . 
20. Excessive drinking. .,,,,... 


-2 
-2 
-1 
-1 
-1 
-1 


4-3 
+ 3 
4-4 
4-2 
4-2 
4-1 


5 
5 
5 
3 
3 
2 



-1 

-1 
-1 
1 


4-1 
4-2 
4-2 
4-1 


o 


21. Not standing up when the "Star 
Spangled Banner" is played. 
22. Being a slacker in time of war . 
23. Neglecting to study your lesson 
24. Making fun of cnpples 
25. Being cross to your brother or sister. 
26. Being a Bolshevik 


-3 

-1 


-M 
+ 1 


-1 

4-1 

-1 




2 
1 
1 
- It 
- 1 
- 1 


-2 


-1 

4-1 
4-1 


-1 




-1 





27. Drinking a great deal of coffee and tea 
28 Whispering in school 


+ 1 
+2 


-2 
-1 


- 3 
- 3 




4-1 




o 
















-734-5 


4-91-5 


4-163-9 


-314-3 


4-36-2 



Average excess and defect scores, and average spans: 
Excess: Female 2.6 Defect: Female .18 Span 
Male .18 Male 32 5 50 



* In Items 1 to 23 the span represents the difference between a female excess and a male 
defect in condemnation, in 24 to 28 between a male excess and female defect, 
t A negative span indicates greater censure by males than by females. 



400 SEX AND PERSONALITY 

The reinforcement of general emotiveness by a specific factor 
appears in the sex divergences in this section, when we compare 
the first five with the last six items in Form A, and again in Form 
B. On the ground of mere emotiveness the items "overworked 
children" (11 in Form A), "a young person totally paralyzed" 
(IS in Form A), and "a young person with a fatal disease" 
(12 in Form B) are surely just as distressing as "a mouse caught 
by a hawk," "a bird with a broken wing," "a wounded deer," or 
"a fish caught on a hook" 1 (items high in the scale). 

But a glance at the last three items adds the specific factor. 
They all instance some weak or delicate creature perceived 
at the moment as helpless and in visible distress. The first three 
items represent a distressing state, in which the helplessness and 
the distress and the weakness of the subject are far less immedi- 
ately perceptible. The female claims to be more moved by the 
sight of the weak and helpless in temporary but palpable distress. 

We note again certain exceptions: "an orphan girl" and "an 
old maid who has always wanted children," although representing 
distressing states rather than instances, appear as Items 1 and 2 on 
Forms A and B. But here the "normal" female excess over the 
male is accentuated by the additional specific factor of sympathy 
with their own sex. "An orphan boy" registers a slightly lower 
score than "an orphan girl" and falls considerably lower in its 
scale. 

Again, objects that are unsightly tend toward a low position 
("a dope fiend," "a toad being swallowed by a snake," "a 
terribly ugly man"), suggesting that pity is counteracted by 
repulsion or disgust. These explanations for what they are 
worth (and they at least offer corroborative evidence) cover all 
the most sex-differentiating items except 2 in Form A "an 
insane person." Why this should seem so much more pitiful to 
females than to males is not clear. Why should female pity 
exceed that of males here more than, say, in the case of " a young 
person with a fatal disease"? 

Summarizing, a comparison of the female and male scores 
suggests or indicates as distinctive sex characteristics: (1) a 

1 "A bee that is drowning" is less maternally appealing bees sting. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 401 



female sympathy (the maternal instinct?) for the weak and help- 
less visibly distressed; (2) a tendency for the female to express 
more pity for female distress; (3) a tendency to more female than 
male sympathy with things attractive in appearance as con- 
trasted with the ugly. 

TABLE 71. SEX DIFFERENCES EN DEGREES OF CENSURE 



M-F test items 


Excess 


Defect 


Span 


3 

Ex- 
tremely 
wicked 



Not 
really 
bad 


Form A* order of female excess over male 
defect: 
1. Boys fighting . . . 


8 


4-7 


is 


3 


+3 


2. Swiping fruit out of orchards 


6 


+ 8 


14 


-3 


+ 3 


3. Shooting rabbits just for fun. . 


4 


+ 8 


12 


2 


+ 3 


4. Boys smoking before they are 21 
5. Indulging in "petting" 
6. Going to bed without saying your 
prayers 


-4 

-5 

-5 


+7 
+6 

+ 5 


11 
11 

10 


-1 
-2 

-2 


+ 3 
+3 

+2 


7. Stealing a ride on a truck 


-6 


4.3 


9 


2 


+ 1 


8. Moderate drinking ,.,,,,., - - , - 


4 


+ 5 


9 


-2 


+ 2 


9. Not brushing your teeth. . . 
10. Having fits of temper .... . . 
11. Boys teasing girls ... 
12. Putting pins on the teacher's chair. 
13. Telling a lie to avoid punishment 
14. Using slang 


-4 
-2 
-3 
-4 
-3 
-3 


+ 5 
+ 7 
+ 5 
+ 3 
+ 3 
+ 3 


9 
9 
8 
7 
6 
6 


-2 

-1 
-1 
-2 
1 


+2 
+3 
+ 2 
+ 1 
+ 1 
+ 1 


15. Picking flowers in a public park. . 
16. Laziness 
17. Boy running away from home 
18. Breaking windows . . . 
19. Insulting the defenseless 
20. Excessive drinking ......... 


-2 
-2 
-1 
-1 
-1 
-1 


+ 3 
+ 3 
+4 
+ 2 
+ 2 
-j-1 


5 
5 
5 
3 
3 
2 



-1 

-1 
-1 
1 


+ 1 
4-2 
+2 
+ 1 


o 


21. Not standing up when the "Star 
Spangled Banner" is played. . 
22. Being a slacker in time of war . . 
23. Neglecting to study your lesson .... 
24. Making fun of cripples 
25. Being cross to your brother or sister. . 
26. Being a Bolshevik ... 


-3 

-1 


+ 1 
+ 1 


-1 

+ 1 

-1 




2 
1 
1 
- It 
- 1 
- 1 


-2 


-1 

+ 1 
+ 1 


-1 




-1 





27. Drinking a great deal of coffee and tea 
28 Whispering in school 


4-1 
+ 2 


-2 
-1 


- 3 
- 3 




+ 1 




o 
















-73 + 5 


+ 91-5 


+ 163-9 


-31 + 3 


+ 36-2 



Average excess and defect scores, and average spans: 
Excess: Female 2 6 Defect: Female 18 Span 
Male .18 Male 32 5 50 



* In Items 1 to 23 the span represents the difference between a female excess and a male 
defect in condemnation, in 24 to 28 between a male excess and female defect, 
t A negative span indicates greater censure by males than by females. 



402 SEX AND PERSONALITY 

TABLE 71. SEX DIFFERENCES IN DEGREES OF CENSURE. (Continued) 



M-F test items 


Excess 


Defect 


Span 


3 
Ex- 

tremely 
wicked 




Not 
really 
bad 


Form Bt order of female excess over male 
defect: 


-4 


4-8 


12 




-j-4 




-5 


4-6 


11 


3 


4-2 


3. " Talking back" to the teacher 

4, Spitting o the sidewalk 


-5 
-5 


4-6 
+ 5 


11 
10 


-2 
-1 


4-2 

4-1 


5. Disobeying a "keep off the grass" sign 
6. Beating your way on a freight train. . . 
7. Stealing candy from the grocery man. . 
8 Being ashamed of your parents .... 


-4 
-3 
-3 
-3 


4-S 
4-6 
+ 5 
4.5 


9 
9 
8 
8 


-1 

-2 
2 


-* r* <s 

-4-4-4-4 


9. Being "yellow" or cowardly 


-2 


4-6 


8 


-1 


4-2 




-3 


4-4 


7 


-1 


4-2 


H. Accepting wrong change when it is in 
your favor 


-1 


+6 


7 





4-2 


12 Not going to Sunday School .... 


-3 


44 


7 


1 






-2 


4-4 


6 





4-2 




-3 


43 


6 


-1 


4-1 


15. Deceiving others without actually 


-3 


4-2 


5 


1 





16. Shooting craps 


-3 


4-2 


5 


-1 


4-1 


17. Girls fighting 





+4 


4 




4-2 


18. Playing cards for money 


-1 


4-2 


3 


1 





19. Discourtesy to people about us 
20. Stealing a dollar from your mother's 
purse 


-2 



4-1 
-2 


3 
2 


-1 

o 



-1 


21. Disobeying your parents 
22. Matching pennies for keeps 


-1 
-2 



+ 1 

4-2 


2 
2 
2 


-1 



o 





o 


24. Being extravagant 





4-2 


2 


o 


4-1 


25. Picking on somebody weaker than 


-1 


-1 





1 





26. Not adding to the collection box at 


-1 


-1 





-1 





27. Stealing a dime from your mother's 


-f 1 





- 1 


4-1 







+ 3 





- 3 


4-2 



















-604-4 


489-4 


149-5 


-234-4 


4-32-1 



Average excess and defect scores, and average span: 

Excess: Female 2 . 1 Defect: Female . 14 Span 

Male .14 Male 3.2 5.1 

Averages for Forms A and B combined: 

Excess: Female 2 . 3 Defect: Female . 16 Span 

Male 16 Male 3.2 5.3 



t In Items 1 to 25 the span represents the difference between a female excess and a male 
defect in condemnation, in Items 26 to 28 between a male excess and a female defect. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 403 



Taken at their face value the scores also indicate a keener 
sense of pity in the female and a corresponding callousness in the 
male. But the true interpretation of the general tendency of 
females to more emphatic expression of feeling requires separate 
discussion. 

Ethical Reactions. Here the subject is presented with a list of 
acts of various degrees of "wickedness" and is asked to grade the 
badness of each on a descending scale of 3, 2, 1, 0. The response 
weights are to be interpreted in the same manner as in the 
preceding sections of this exercise. For example, "stealing a 
ride on a truck" has the response weights 2, 2, +1, +1 for 
the 3, 2, 1, degrees of badness respectively, showing that 
females more often than males check the 3 or 2, and males more 
often than females the 1 or 0. 

The results are summarized in Table 71 in which the "excess" 
score is arrived at by doubling the weight for "3" (extremely 
wicked) and adding it to the weight for "2 " (decidedly bad), the 
"defect" score by doubling the weight for "0" (not really bad) 
and adding it to the score for "1 " (rather bad). Table 71 gives 
separately the weights for "3 " and "0" as well as the "excess" 
and "defect" scores. Table 72 gives the male and female aver- 
ages for each degree of censure. 

TABLE 72. AVERAGE SCORE WEIGHTS FOR EACH DEGREE OF CENSURE 

(Numbers in parentheses refer to the number of items in which a higher 

proportion of males or females recorded the particular degree of 

censure in question) 



Degrees 


Extremely 
wicked 

3 


Decidedly 
bad 

2 


Rather 
bad 

1 


Not really 
bad 



Form A 
General averages 


1.2 


1 


.9 


1.4 


Male average 


(3) .1 


(7) .3 


(19) .8 


(18)1.3 


Female average 


(19)1.1 


(14) .7 


(3) .14 


(2) .07 


FormB 
General averages 


.9 


1.2 


1 


1.2 


Male average 


(3) .15 


(7) .25 


(14) .9 


(14)1.1 


Female average 


(18) .8 


(16) .7 


(2) .07 


(1) .04 













404 SEX AND PERSONALITY 

A glance at Table 71 reveals the greater tendency of females to 
excessive censure, as most items show a span of female excess over 
male defect, few the reverse. Table 72 indicates this difference 
succinctly. The average scores under each degree of wickedness 
suggest that the items seem to the subjects as a whole fairly 
representative of all degrees of badness, since the differences in the 
general average scores under each degree are not very marked, 
the most lenient judgment ("not really bad") being slightly the 
commonest. But the division into female and male scores shows 
that this conclusion is delusive, for the result is reached by a 
female tendency to severity of judgment being balanced by a 
male tendency to leniency. To males more of the offenses listed 
are trivial, and to females serious. Statistically, the female 
"extremely wicked" judgments average 11 and 5 times more 
severe than those of the males (for the two forms respectively), 
and the male extremely lenient judgments ("not really bad") 
18 and 27 times more lenient than those of the females. We may 
not conclude that females are really morally more sensitive or 
more particular than males, since the same tendency to excess of 
expression in the female characterizes the responses for the four 
emotional sections, including the emotion of anger; and who would 
claim that females are more pugnacious or more irascible than 
males are? Moreover, had females been really sterner moralists 
than males to the extent of the difference in the scores this would 
have forced itself upon general notice long ago. 

But perhaps a closer analysis will reveal specific factors as well 
as this sex difference in general emphasis. Two questions suggest 
themselves : (1) Does the female severity apply more to the lighter 
or the graver offenses? (2) Is a given sex more condemnatory of 
offenses which characterize the opposite sex? 

The first question may be answered by comparing the scores 
for the most, with those for the least, serious offenses. Although 
there is not, of course, any one true order of seriousness for this or 
any other list of offenses, we have from each form selected for 
comparison three which seem to us among the most reprehensible 
and three among the least so. The "excess" and "defect" 
scores for these are given in Table 73, which shows a marked 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 405 



tendency for the female sex to be harsher than the male in judging 
petty offenses, but to be relatively less harsh in the case of more 
serious offenses. 

TABLE 73. SEX DIFFERENCES IN THE CENSURE OF SERIOUS AND TRIVIAL 

OFFENSES 



M-F test items 


Excess 


Defect 


Span 


Form A 
Serious offenses: 
Excessive drinking 


-1 


4-1 


2 


Insulting the defenseless ... 
Making fun of cripples. . . 


-1 



+2 
1 


- 3 
1 


Average difference between female ex- 
cess and male defect of severity is 1 3 . 
Trivial offenses: 
Not brushing your teeth 
Whispering in school 
Usinsc slang 


-4 

+2 
_3 


+5 
-1 

4-3 


9 
3 
6 


Average difference between female ex- 
cess and male defect is 4. 
FormB 
Serious offenses: 
Stealing a dime from your mother's purse. 
Picking on somebody weaker than yourself 
Cheating in examination 


+1 
-1 




-1 

4-2 


2 

2 


Average difference between female ex- 
cess and male defect is 3. 
Trivial offenses: 
Losing your temper 
Being extravagant 


-3 



+3 
+2 


6 
2 


Spitting on the sidewalk 


-5 


4-5 


10 


Average difference between female ex- 
cess and male defect is 6. 









We come now to our second question, whether a sex tends to 
condemn the more the offenses of the opposite sex. We may 
divide the items into four groups of offenses: distinctively male, 
distinctively female, common, and ambiguous. By "common" 
is meant those committed about equally by both sexes, and by 
"ambiguous" those about which difference of opinion might be 
expected, but which are not so obviously special to one sex as to 
justify their inclusion in our first two classes. Table 74 gives the 
data according to our own classification of the items. 



406 



SEX AND PERSONALITY 



TABLE 74. CENSURE SCORES OP MALE AND FEMALE OFFENSES 



M-F test items 


Excess 


Defect 


Span 


Distinctively male offenses: 
1. Boys fighting . . . . . 


- 8 
- 6 

- 1 

- 5 

- 5 

- 6 
- 1 


+ 7 

-- 3 
-- 2 

+ 5 
+ 6 
+ 6 

i! 

+ 4 


15 
12 
9 

3 

'f 

11 

11 
14 


2. Shooting rabbits just for fun 


3. Stealing a ride on a truck . . . . . 


4. Boys teasing girls 


5. Breaking winnows 




7. Spitting on the sidewalk..;. . . . .. .. 
8. Beating your way on a freight train . . . ... 
9. Swearing 


10. Picking on somebody weaker than yourself . 
) 1 , Boys smoking bef 01*$ they are 21. . . - . - .... 


12. Swiping fruit out of orchards 


Average span of female excess (109 + 13). 

Distinctively female offenses: 
1. Girls fighting. . . .... 
2. Drinking a great deal of coffee and tea 
3. Being extravagant ... 


-48 


+ i 


+62- 1 

J 

+ 2 


109 
8.4 

4 
~~2 


Average span of female excess (6 3 -f- 3 ) . . 

Offenses common to both sexes: 
1. Indulging in "petting". 
2. Going to bed without saying your prayers 
3. Not brushing your teeth ... 
4. Telling a lie to avoid punishment ... . 
5. Picking flowers in a public park 


1 I++I 1 1 1 1 1 I++I 1 1 1 1 1 1 + 


+ 6-2 

n 

+ l 



+ 5 

i! 
\ 

\ 



+ 2 
+ 3 


+6-3 
1 

11 
10 
9 
6 

5 
1 

~8 
6 
3 
~2 


-1 

6 


7. Neglecting to study your lesson 
8. Being cross to your brother or sister. 
9. Whispering in school . 


10. Being ashamed of your parents .... 
1 1 . Accepting wrong change when it is in your favor 


13. Deceiving others without actually telling a he 
14. Discourtesy to people about us 
15. Stealing a dollar from your mother's purse . . 
16. Disobeying your parents 


17. Cheating in examination 
18. Not adding to the collection box at church . . 
19. Stealing a dime from your mother's purse . . 
20. Being an atheist 






Average span of female excess (87 10 -* 22) 

Ambiguous offenses: 
1 Moderate drinking . - - 


l + l 1 1 1 1 1 1 1 1 1 1 III 1 

+ 

^1 


+ 50- 4 

f! 

- 1 
- 1 

+ 1 

II 

+ 6 

+ 4 
+ 4 

+ 0* 
+ 8 

+ l 

+ 2 


87- 10 
3 5 

9 
9 

2 
-1 

11 
9 
8 
8 

7 
6 

12 
-1 


2. Having fits of temper 


3 Putting pins on the teacher's chair 


4. Not standing up when the "Star Spangled Banner " 




6 Being a slacker in time of war .... 


7. " Talking back" to the teacher 


8. Disobeying a "keep off the grass" sign . ... 
9. Stealing candy from the grocery man 
10. Being yellow" or cowardly 


12. Not going to Sunday School 






15. Playing marbles for keeps 




17. Being a Bolshevik 


18. Insulting the defenseless 


Average span of female excess (106 2 4- 18). . . 


-45 + 1 


+62 - 2 


106-2 
5.8 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 407 



Comparing the average spans of female excess over male defect 
in severity, 8.4 for specifically male as against 3.5 and 5.8 for the 
common and ambiguous offenses, and 1.00 for the three specifi- 
cally female offenses, the statistics evidence a female tendency 
to be more severe than males upon male offenses, and if we may 
take three items as enough to judge by, to be more lenient toward 
female offenses than toward offenses common to both sexes. 

An interesting instance of the effect upon the scores of the 
combination of the general female tendency to excessive expres- 
sion with the tendency of a sex to condemn the other sex's offenses 
more severely is afforded by the figures for the parallel items 
"boys fighting" and "girls fighting." 



M-F test items 


Excess 


Defect 


Span 


Boys fighting 


-8 


+7 


15 


Girls fighting 





+4 


4 











The combination of the two female tendencies to severity in 
general, and to condemn male offenses in particular, makes the 
span for "boys fighting" the highest among all the 56 items on 
both lists; while on the other hand this tendency to condemn in 
general is in the case of the specifically female offense reduced to 
insignificance in "girls fighting" by the sex's condonation of its 
own sex's offense. 

We may now ask whether or how far it is ethical discrimination 
that the responses are revealing. To what extent have the 
specified types of behavior been judged according to moral 
standards only? 

One way of approaching this question is to have the items 
ranked by independent moral experts (presumably a chosen few) 
and then to see of which sex the scores most nearly correspond to 
their ranking. We might conclude that, other things equal, the 
less consonant sex admits the more emotional bias; its judgment 
is more an emotional and less a purely ethical judgment. But 
this would entail an additional research. And in any case the 
chosen experts might be expert only in the social morality of their 



408 



SEX AND PERSONALITY 



TABLE 75. CENSURE SCORES OF FOUR GRADES OP OFFENSES 



M-F test items 


Excess 


Defect 


Span 


Grade A (the most serious offenses) : 
1. Insulting the defenseless 


-1 


+2 


3 


2i Kxcessive drinking M 1 . .... . . , . . . 


-1 


+1 


2 


3. Making fun of cripples 





1 


-1 


4- Picking on s^tocbody weaker than yourself . . , 


-1 


-1 













Average 


-0.75 


+0.25 


1 


Grade B (serious but not so grave as A) : 
1. Lying to avoid punishment 


-3 


+3 


6 


2. Accepting wrong change when it is in your favor 
3. Stealing a dollar from your mother's purse 
4. Cheating in examination 


-1 




+6 
-2 
+2 


7 
-2 
2 


5. Deceiving others without actually telling a lie 
6. Swearing M 


-3 
-5 


+2 
+6 


5 
11 


7. Boys teasing girls M 


-3 


+5 


8 


8. Girls fighting F 1 





4-4 


4 










Average 


-1.8 


+3.2 


5.1 


Grade C: 
1. "Talking back" to the teacher 
2. Spitting on the sidewalk M 


-5 
-5 


+6 

+5 


11 
10 


3. Losing your temper 


-3 


+3 


6 


4. Being extravagant F 





+2 


2 


5. Boys fighting M 


-8 


+7 


15 


6. Beating your way on a freight train M . . . . 
7. Neglecting to study your lesson 


-3 
-1 


+6 

o 


9 
1 


8. Boys smoking before they are 21 M... 


4 


+7 


11 


9. Breaking windows M 


-1 


+2 


3 


10. Whispering in school 


+2 


-1 


-3 


11. Disobeying a "keep off the grass" sign 


-4 


+5 


9 


Average 


-2.9 


+3.9 


6.7 


Grade D (the least iniquitous) : 
1. Being an atheist 


+3 





-3 


2. Not adding to the collection box at church 


-1 


-1 





3. Being a Bolshevik 


+1 





-1 


4. Not brushing your teeth 


-4 


+5 


9 


5. Moderate drinking 


-4 


+5 


9 


6. Using slang 


-3 


+3 


6 


7, Playing cards for money. . , , . 


-1 


+2 


3 










Average 


-1.3 


+2 


3.3 











1 M and F indicate that the item is considered a specifically male or female offense. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 409 

time and place, freedom from which in an unsophisticated child 
(as indeed in an adult also) might evidence a more genuinely 
ethical judgment. 

The only plan immediately practicable is to take the compared 
sex responses as they stand and, after ranking to one's own satis- 
faction the various items in a scale of offensiveness, to see how far 
one sex shows the more severe judgment than the other for the 
offenses judged by us more serious, and conversely. For reasons 
given this method is precarious, but we can with the material at 
hand apply it. The 30 items which seemed easiest to place at 
one of four levels of badness are ranked in Table 75. 

If our ranking represents a more correct moral judgment than 
those of the subjects themselves, the analysis shows, after allow- 
ing for the effect of specific sex bias: 

1. A tendency, except for morally colorless practices enumer- 
ated in Grade D, for the male judgments to be nearest the female 
judgments in severity as the offenses increase in gravity. 
^2. That as compared with males, females are particularly 
severe on minor offenses which are yet just sufficiently offensive 
to merit condemnation; but 

3. That when the "offense" becomes so inconsiderable as to be 
very questionably an offense at all, even the females are apt to 
relax their proportionate severity a little. But again 

4. That even in the morally indifferent items (Grade D) there 
are still some against which the distinctive female severity is 
marked ("using slang," "not brushing your teeth," and "moder- 
ate drinking"). 

The results support the view that the male has on the whole a 
more "objective" moral judgment than the female, who tends to 
exaggerate minor offenses. 

To the question then whether it is ethical discrimination 
(capacity for reasoned objective valuation) that is being meas- 
ured, apart from the dependence of the answer upon the particular 
sentiment tapped by the particular item, and apart from our 
general human incapacity to keep our judgments clear of bias, our 
analysis suggests the answer that the female judgments tend, on 
the whole, to be more emotional and less objective than the male. 



410 



SEX AND PERSONALITY 



This view is supported by such curiosities as the markedly distinc- 
tive female condemnation of trivial offenses like "not brushing 
your teeth " and "using slang." 

Preference for Alternatives. The subject is presented with 
pairs of situations, occupations, or objects, and records against 
each pair which of the two, if either, he likes the better. 

In Table 76 the items are arranged in descending order of male 
and female preference (for one rather than for the other alter- 
native in each case). An example will explain the symbols used : 

TABLE 76. PREFERENCE ORDER FOR CHOICE OF ALTERNATIVES 

Items in descending order of male and female preference 

Forms A and B combined 



Male preference 




Span 


Female preferenc 


e 


yi/ Work with men v. 1 .... 
^)Give a report verbally v. 


+22 
+ 5 


35 
or 
37 

10 


Jr* Work with women. . 

^)Give a report in writ- 
ing 


- 5 


3. Camping v. 7 


+ 4 


7 


3. A poorly cooked meal, 




4. A well-cooked meal, but 
soiled table linen v. 3 
5. Command others v. 4 . . 
/cy An auto with scruffy 


+ 4 
+ 4 


9 
9 


but linen spotless. . . 
4. Persuade others. . . . 

5. Work of a single kind 
6. Book work 


- 5 
- 5 

- 4 

- 4 


paint but excellent 
c*^~ motor v. 8 


+ ^ 


6 






(l) Great variety of work v. 
5 ... 


+ 3 


7 


7. Living in good hotels.. 


- 3 


8. Laboratory work v. 6 . . 


+ 3 


7 


6 An auto with fresh 
^ paint but only fairly 
good motor , 


i 


(*. Work involving many 
details v. 12 


+ ? 


4 


JMInteresting work with 
small income 


- 3 


tf{D Uninteresting work with 
,^-s large income v. 9 
l4 Make plans v. 14 


+ 2 
+ 1 


5 


10. Work in one location 
^Live in the citv 


- 2 
- 2 


jX Live in the country v. 1 1 


+ 1 


3 


^ v "* v y 
Work involving few 

details 


- 2 


13. Change from place to 
place v. 10 


+ i 


3 


13. Work under a respect- 
ed superior .... 


- 2 


14. Work for yourself v. 13 . 


+ 1 


3 


(yJ Carry out plans ... . 


- 1 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 411 

in Item 3 on the male preference side of the table, +4 means that 
a markedly higher proportion of males than of females say they 
prefer "camping." "V.7" indicates the alternative, given in 
Item 7 on the female preference side, to which they prefer it, 
namely, "living in good hotels." The 7 under the heading 
"Span" represents the span or sex divergence in preference got 
by summing the male preference score for "camping" and the 
female preference score for its alternative, "living in good 
hotels." (One item, "work with women vs. work with men" 
occurred in both Forms A and B with female preference scores of 
13 and 15 respectively, and male scores of 22 for "work with 
men" in both cases.) 

On the basis of conclusions from our previous exercises pre- 
dictions were ventured, before consulting the scores, as to which 
sex would score higher for each alternative. The predictions for 
Items 1, 3, 4, S, 6, and 8 on the male preference side of the table 
and their corresponding alternatives, 1, 7, 3, 4, 8, and 6 on the 
female preference side proved correct: sex traits now familiar 
explain the sex differences in response. The sex-to-sex tendency 
(M 1 and F I), 1 the male out-of-door interest and love of adven- 
ture (M 3 and F 7), the female interest in decorativeness and 
beauty of appearance (M 4 and F 3 and M 6 and F 4), the greater 
male interest in the mechanical, and perhaps the female interest 
in some kinds of literature (M 8 with F 6) suffice or help to explain 
all the more significant contrasts in which prediction was verified. 

The contrast "work with men" and "work with women," 
appearing in both forms, has the singularly high sex divergence 
span of 37 and 35, the next highest in the section being 10. But 
in both cases the item "work with men " has a plus index of 22, its 
counterpart "work with women" having a minus index of 15 
and 13. This indicates that though each sex evinces strongly 
the sex-to-sex tendency, at the same time more females express a 
preference for male than males for female coworkers. 

The items in which predictions failed run as follows: 

1 "M 1" means Item 1 on the male preference side, "F 1" means Item 1 on 
the female preference side. 



412 SEX AND PERSONALITY 



M-F test items 



Scores 



Span 



M F 

1. Give a report verbally vs. report in writing 

M F 

2. Great variety of work vs. work of a single kind 

M F 

3 . Work involving many details vs. work involving few details 

M F 

4. Uninteresting work with large income vs. interesting work 



with small income. 



M F 
5. Change from place to place vs. work in one location 



+5-5- 
+3-4 = 
+2-2 = 



+2-3 
+1-2 



10 



7 



4 



5 



3 



Predictions in these cases were rash and precarious because 
the sex differences recorded are not really or clearly derivable 
from the distinctive traits disclosed by previous exercises, but 
from other characteristics not contained in them and not so far 
revealed by our inquiry. 

Summarizing: (1) most cases of widest sex span support, and 
none contradicts, conclusions as to sex differences ventured 
elsewhere; but (2) the form of the exercise admits of so many 
variant reactions under the same recorded scores as to make 
generalizations from the narrower spans precarious. 

Sex Differences in Emotional Expressiveness. The tendency 
of females to more extreme responses than males has already been 
noticed, specifically in anger responses and moral censure. We 
may now consider this question more generally. For all four 
emotions and for ethical judgment we have summarized the 
evidence of a distinctive female tendency toward excess, and a 
distinctive male tendency toward defect, of expressed emotion 
and censure. The table at the top of page 413 gives for Form 
A and Form B combined, and separately for anger, fear, 
disgust, pity, and wickedness, the number of cases in which 
the highest or lowest degrees of feeling were expressed by each 
of the sexes. That is, there are 151 items showing female excess 
as compared with 9 showing male excess, but only 13 cases of 
female defect as compared with 142 of male defect. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 413 



Emotion 


Highest degree 


Lowest degree 


Male 


Female 


Male 


Female 


Anger 


1 
2 


6 


24 
29 
35 
26 

37 


19 
33 
32 
23 

35 


5 
4 
1 


3 


Fear 


Disgust 


Pity 


Censure 


Total . ... 


9 


151 


142 


13 



Moreover, the average score weight in the cases of male excess 
and female defect were much lower than in the items showing 
male defect and female excess. These are as follows, for Form A 
and Form B combined: 



Emotion 


Excess 


Defect 


Anger: 
Female 


3.9 


.3 


Male.. 


.06 


1.8 


Fear: 






Female 


4.2 


.25 


Male 


.1 


5 2 


Disgust: 
Female 


7.7 


17 


Male.. 


.0 


5 4 


Pity: 
Female 


4.8 


.04 


Male 


.0 


3.6 


Censure: 






Female. . . 


1.9 


11 


Male 


.24 


2.5 



The above figures can be summarized by saying that the female 
tends much more than the male to extreme expression in all four 
emotions and in moral censure, particularly in disgust, rather less 
in pity, less still in fear, and least in anger and censure; though her 
excess, if less marked for censure than for pity, fear, and anger, is 
more ubiquitous. 



414 SEX AND PERSONALITY 

How are we to interpret these figures? In regard to the four 
emotions, two questions await answer: "Why is the female gener- 
ally more expressive than the male?" and "Why is her excess 
over the male least in the case of anger?" To the second question 
it is easy to suggest an answer in terms of social expectation, 
training, and bias, and to a certain extent in terms of innate 
factors: the girl is socially trained to express fastidiousness, and 
permitted to express fear more freely than the boy, and in certain 
directions to express more pity, but not to show more anger. 
Again, the female may be innately more prone to fear and active 
sympathy, and both more especially trained and perhaps more 
prompted by innate sex motives to shun the ugly and unsightly, 
but not to express more anger. It is the girl rather than the boy 
from whom society exacts the sweeter temper and expressions 
not of anger but of gentleness. Hence we can ascribe the com- 
parative mildness of the female in the case of anger to innate or to 
culture factors, or to both. 

But it remains to consider the first of the two questions: "Why 
does the female still show herself more angry than the male?" and 
also to compare the results for moral censure with those for the 
emotions. 

Three possibilities suggest themselves: 

(1) The female may be really the more emotional sex; she 
" feels things more" than the male does; or 

(2) a greater freedom of expression may be socially encouraged 
or sanctioned in the female; or again, 

(3) her training and occupations promote in her a greater 
tendency to verbal exaggeration; or 

(4) two or more of these may combine. 

The first hypothesis cannot be directly tested; the only objec- 
tive test is through expression, but whether expression corresponds 
with that which it expresses is the very point at issue. But there 
are reasons for mistrusting the adequacy of this hypothesis. 
Though psychologists might agree that the innate impulses of 
flight, repulsion, active sympathy, are stronger in the female sex, 
with their accompanying asthenic emotions of fear, disgust, and 
pity, surely no one maintains the same of pugnacity with the 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 415 

asthenic emotion of anger. Though an innate factor a genu- 
inely keener emotion may therefore account for some of the 
excess expression in fear, disgust, and pity, a residuum of excess 
over the male, represented by the excess figures for anger, remains 
to be accounted for in all the sections; and this cannot acceptably 
be referred to greater intensity of feeling. If then this residuum 
can be adequately covered by one of the other hypotheses we shall 
not be justified in referring it to the first. 

One inclines rather to seek an explanation from the two 
environmental factors. On the one hand, in Western society 
at any rate, the boy is tacitly and openly expected to be more 
"Spartan" than the girl. Crying in a girl is condoned on 
occasions when in a boy it would be voted " sissy." No boy may 
be "as timid as a girl." To be overnice about dress or deport- 
ment is considered effeminate. Control over feeling, or at least 
over its expression, is considered manly in a man but not 
womanly in a woman. Emotional display is socially expected in 
a woman (the man rather likes it in her), but emotional control 
is expected in the man (the woman rather admires him 
for it). 

Moreover, as the sexes mature their customary social milieus 
increasingly differ, the boy taking to business and the girl to 
domestic and social life. Business keeps feeling in the back- 
ground, social gatherings bring it out. Emotional exuberance 
befits the one situation, but reticence the other. Exuberance 
easily passes into exaggeration, and the girl forms the habit of 
saying more than she means, while with boys the tendency is the 
other way. The conclusion may be ventured that a cultivated 
masculine reticence and feminine exuberance combine to account 
for the scores. On these assumptions it becomes unnecessary to 
resort to our unverifiable "greater emotionality" hypothesis. 

Lastly, it is easier to explain the sex divergence in severity of 
moral judgment by the environmental hypothesis. It indicates 
why ethical judgments share with anger the lowest place on the 
sex divergence scale. With fear, disgust, and pity, special 
reasons have been found for the additional female excessiveness 
beyond a general residuum. That general residuum has been 



SEX AND PERSONALITY 

-attributed to male inhibition on the one side and to female 
exaggeration on the other. 

EXERCISE 5: INTERESTS, LIKES, AND DISLIKES 

Exercise 5 is divided into eight series of likes and dislikes, in the 
first five of which the subject has to declare whether he likes, 
dislikes, or is indifferent to 

A. Vocational occupations; 

B. People with certain characteristics; 

C and D. Certain forms of pastime, entertainment, occupa- 
tions, or situations; 

E. Certain books. 

The last three series ask him which items from a given program 
he would like, dislike, or neither like nor dislike 

F. To draw (if an artist) ; 

G. To report on (if a newspaper reporter) ; 
H. To do or visit (if on a period of travel). 

If the information exercise is an indirect, this one is a direct, 
inquiry into individual interests, and to that extent would appear 
to obtain its results less from the accidents of experience and more 
from the emotional dispositions of the subject, and so to reflect 
more directly his actual temperament and personal habits of 
mind. By actual we do not necessarily mean innate, for our 
likes and dislikes are also influenced by environment. 

Our discussion is based upon data from 158 subjects in the 
grade school, 200 in high school, 52 in college, and 80 nonacademic 
adults ; 590 in all. Each group has an equal number of males and 
females. 

Predictive Analysis. For the first section in Form A (on occu- 
pation preferences) and for 72 cases in Form B where a prefer- 
ential sex like or dislike seemed deducible from the conclusions of 
our analysis of Exercise 3, predictions were ventured as to which 
sex would score higher under "L" (like), and "D" (dislike). If 
tentative inductions as to the emotional dispositions and conse- 
quent directions of interest of the two sexes derived from an 
investigation of a situation where they were only indirectly 
exhibited, and might be expected to have been obscured by other 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 417 



circumstances, were confirmed by an immediate inquiry into these 
very interests themselves, this would mean that the traits were so 
dominant that, even in a situation where environment and experi- 
ence (in the shape of factual information) were given every chance 
of determining the responses, these sex differences in emotional 
tendencies and interests had proved the one set of differences that 
clearly emerged all through. 

Out of the 48 predictions of L and D sex predominances ven- 
tured for Section A, Form A, 41 instead of a chance 16 proved 
correct. But in venturing a prediction one had to remember: 
(1) that the question at issue was not of the absolute amount of 
like or dislike for a given occupation, but of which sex liked or 
disliked it more, a pronounced sex preference being perfectly 
compatible with an absolute dislike on the part of the sex express- 
ing the comparative preference; (2) that even a very small 
difference in the sex responses would produce a score for one or 
other sex, though so small as to be relatively unreliable. 

For both these reasons several erroneous forecasts might have 
been anticipated. That 41 instead of the chance 16 proved 
correct to this extent supported the hypothesis upon which the 
predictions were based. The 7 incorrect forecasts concerned four 
occupations for all of which predictions had been diffident, 
through inability to estimate or balance conflicting factors. 
These were as follows: 



Item 


Forecast 


Actual 
scored result 


Preferential 
index 


Architect 


L -,/> + 


L+,D - 


3 


Chef or cook 


D + 


DO 


3 


Journalist 


L+,D- 


L-.D + 


6 


Draftsman 


L-,D + 


L+.D 


10 











It might be added that for two of these four the preference index 
(or degree of sex divergence) is the lowest for the section, making 
prediction the more precarious. Of the 25 items on the list, in 9 
the forecast had been marked as diffident or dubious; 4 proved 
incorrect and 5 correct. The correlation of confidence with 



418 



SEX AND PERSONALITY 



correctness, and diffidence with half correctness, supports the 
hypotheses which were the bases of the forecasts. 

Of the 72 cases in the various sections of Form B for which 
prediction as to sex preferences was hazarded on the basis of our 
assumed sex differences, 68 proved correct. 

We may now pass to the a posteriori analysis. 

Relative Differentiating Power of the Various Sections of the 
Exercise. Of the eight sections it seemed that A (on preferred 
occupations) would most repay study. As it turned out, this 
section shows the largest sex difference, and B the least. Table 77 
gives our statistical data on the relative amount of sex difference 
brought out by the eight sections. The figures of Table 77 
are reached by amalgamating the two forms, A and B, and 
averaging the score weights for L, D, and N respectively for each 
section. The sum of the L and D averages yields one, and the 
average N score weight another, index of the degree to which the 
responses divide the sexes. A high " affective span " and to a less 
degree a low "hesitance" (neutral) score indicate marked sex 
divergence, and vice versa. The order is that of decreasing 
significance, as determined by the affective span. 

An example will make the method clear. Section A contains 
SO items in all, 25 in each form. The average score for L, taking 
female and male scores together, is 5.4, and the average D or 
dislike score, 4.7, yielding a difference between them, or affective 
span, of 10.1. The average score for "neither like nor dislike" 
is 1.4. We call this the "hesitance index." A high affective 

TABLE 77. DISCRIMINATIVE VALUE OF THE SEVERAL PARTS OF EXERCISE 5 



Section 


No. of 
items 


Average 
L 


Average 
D 


Affective 
span 


Average 

N 


Hesitance 
index rank 


A 


SO 


5.4 


4.7 


10.1 


1.4 


(3) 


C 


27 


4.9 


2.9 


7.8 


2.4 


(8) 


G 


12 


3.9 


3.2 


7.1 


.9 


(D 


H 


22 


4.2 


2.8 


7.0 


1.5 


(4) 


D 


40 


3.8 


3 


6 8 


1.38 


(6) 


F 


16 


3.9 


2.8 


6.7 


1.6 


(4) 


E 


46 


4.7 


1.7 


6.4 


1.3 


(2) 


B 


24 


1.9 


2.2 


4.1 


2 


(7) 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 419 

span and a low hesitance index shows that the items differentiate 
the sexes markedly; a low span and a high hesitance index shows 
that the items are relatively nondifferentiating. 

Section A : Occupations. In interpreting the data on this 
section it is well to remember that in default of actual experience 
with given occupations the subjects' responses are likely to be 
influenced largely by cultural sex bias; especially the younger and 
less intelligent subjects will be unable to dissociate their sentiment 
toward an occupation from its inherent interest to them. Even 
in occupations fairly equally open to both sexes a bias may appear ; 
for example, the "he-man" may by his very tendency to be self- 
assertive affect a distaste for an occupation which women are 
rapidly taking up. In short, the sex differences that this section 



TABLE 78. PREFERENCE INDICES OP OCCUPATIONS BY SEX 
Form A 



Occupation 


Male 
preference 
index 


N 


Forest ranger 


-f IS 





Building contractor 


+ 11 





Stock breeder 


+ 11 


+ 3 


Draftsman . . .... 


+ 10 


-1 


Auto racer 


+ 9 


+ 1 


Soldier 


+ 8 





Dairyman 


+ 7 


+2 


Detective 


+ 6 


+ 1 


Architect 


+ 3 


II 


Average 


8.8 




Occupation 


Female 
preference 
index 


N 


Dressmaker 


-25 


+ 1 


Librarian . . . 


-22 


+ } 


Nurse 


-18 





Florist . . . 


-16 


+3 


Private secretary 
Singer 


-16 
-14 


* 


Music teacher 


-13 


if 


Social worker 
Bookkeeper 


-12 
-10 




+2 


Novelist" 


- 8 





Artist 


- 7 


+ 3 




6 





Clerk in a store 


6 


o 


Preacher . 


- 5 


_2 


Optician 






Chef or cook 


- 3 


+ 3 


Average 


11.6 











420 



SEX AND PERSONALITY 



TABLE 78. PREFERENCE INDICES OP OCCUPATIONS BY SEX. (Continued) 

FormB 



Occupation 


Male 
pref eren ce 
index 


N 


Auto mechanic 


+ 16 


+ 1 


Factory manager 


+ 12 


+ ? 


Aviator ". 


+ 9 


+ 1 


Fisherman 


+ 8 


+? 


Plumber 


+ 8 


+ ? 


Miner 


+ 6 


+ ? 


Explorer 


+ s 





Farmer or rancher 


+ 4 





Average 


8 5 










Occupation 


Female 
preference 
index 


N 


Interior decorator 
Teacher 


-22 
-16 


t? 


Stenographer ... 
Barber or hairdresser 
Waiter or waitress . . . 
Jeweler 


-IS 
-14 
-13 
-12 


+ 3 



+ 2 


Poet 


-10 


+ ? 


Sculptor 


-10 


+ 1 


Magazine illustrator . 


- 9 
- 9 


+0 


Office cleric 


- 9 


o 


Furniture dealer 


- 8 


-1 




- 8 


+ 2 


Landscape gardener 
Orchestra conductor 
Cashier .... 


- 6 
- 5 
- 4 




+ 3 


Research worker ... 
Average ... 


10.2 


-1 



reveals may be acquired no less than native, and the factor of 
"essential sexuality" remains obscure. 

Table 78 summarizes the item weights of the various items in a 
form suitable for exhibiting sex contrast; namely, in two state- 
ments, one of occupations preferred by a higher proportion of 
females, the other of those preferred by a higher proportion of 
males, in orders of decreasing degree of preference. It must be 
premised, in explanation of the table, that a high sex record under 
L (like) predisposes to a high opposite sex record under D (dislike), 
and conversely, since if many of one sex record a positive liking 
for an occupation, there remain fewer of that sex but more of the 
other to record a dislike. Not that the equation will be exact, 
since there is also a third choice, N (for indecision), which may 
attract unequal numbers from the sex residues. Therefore, in 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 421 

establishing the total preference of one sex over the other in a 
given case, account must be taken of both the L and the D 
records by treating the degree of liking expressed by sex A plus 
the degree of dislike expressed by sex B as an index of A's prefer- 
ence over B's for the occupation in question. 

The "preference index" of an item is found by adding the D 
weight to the L weight and attaching the sign ( or +) for L 
to the total, thus making it an index of liking. For example, 
Item 12 in Form A: 

12. Dressmaker -131 +12D +1N 

yields a female preference index of 25. The number against N 
yields an index of hesitance; it registers the relative indecision of 
sex choice, a plus registering a higher proportion of males, a 
minus a higher proportion of females as hesitant or undecided. 
The preference index records the degree, and an opposite sign 
in the hesitance index the decisiveness, of the relative preference of 
a given sex for the item concerned. For instance, a +3N 
signifies that far more males than females were undecided in their 
attitude. An opposite sign to 3 may thus be considered a deci- 
sive, and of 4 or 5 a very decisive, preference. 

Let us consider first the data for Form A. A comparison of 
male with female preferential items brings out the following 
contrasts: 

1 . The most indoor occupations fall on to the female preference 
side. Perhaps "draftsman" and "architect" are exceptions. 

2. Not one of the female preferences can be considered prima- 
rily an outdoor occupation. "Social worker" and "journalist" 
come nearest to exceptions. 

3. All the artistic and decorative occupations are on the female 
side. 

4. The six occupations on the list that involve direct ministra- 
tion to individual comfort, convenience, or welfare are all on the 
female side (cook, social worker, nurse, dressmaker, private 
secretary, and preacher). 

5. All the occupations on the male preference list are also occu- 
pations pursued exclusively or predominantly by males, more so 



422 SEX AND PERSONALITY 

with those having the highest preference indices (forest ranger, 
building contractor, stock breeder), and less so with those having 
the lowest (detective, architect). 

6. On the female side most occupations are occupations pur- 
sued by both sexes, none is a male monopoly; one only, or, if we 
include "nurse," two are female monopolies, and those two have 
nearly the highest female preferences. 

7. The female exceed the male preferences both in number 
(16 against 9) and in average degree of preference (11.6 against 
8.8). 

Turning to the data for Form B, we find all seven points are 
corroborated. As regards the seventh, the female exceeded the 
male preference both in number (17 against 8) and in average 
degree of preference (10.2 against 8.S). 

There are no items that indicate occupations carried on in the 
home, except perhaps "cook," "nurse," and "private secretary" 
(and these are female preferences), so that interest in that 
category is not ascertainable from this section of the exercise. 
We may add two more generalizations applicable to this section. 

8. All the items associated with adventure and bodily risk, 
except missionary, 1 fall within the male preferences (auto racer, 
soldier, detective, miner, explorer, forest ranger, and stock 
breeder). 

9. All the items requiring muscular strength or prolonged 
bodily exercise fall on to the male side (forest ranger, soldier, 
building contractor, stock breeder, auto mechanic, fisherman, 
plumber, explorer, rancher, miner). 

Contrasting for both forms the items that are most preferred 
by males and females respectively (namely, those near the top of 
the tables), and comparing them again with those that are less 
strikingly preferred (those nearer the bottom), we find that voca- 
tions representing a combination of artistic and indoor occupa- 
tions (interior decorator and dressmaker) head the female list 
on both forms, and a more outdoor occupation concerned with 
machinery heads the male list for Form B ; and an active outdoor 

1 The term may suggest to a schoolchild religious ministration and sentiment 
as much as adventure and bodily risk. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 423 

occupation the list for Form A. In general, the occupations 
that are most sedentary tend toward the upper half of the female 
lists. 

Summarizing: the records for Section A suggest that males 
have a favorable, and females a less favorable, sentiment toward 
outdoor vigorous occupations and occupations involving bodily 
risk and adventure; females have a favorable, and males a less 
favorable, sentiment toward artistic and decorative occupations 
and occupations involving immediate care of persons. We may 
name these the "ministrative" occupations to indicate the 
probable point of emotional appeal. The results of our scrutiny 
of this section support our main hypothesis: we find the male 
professedly the more aggressive and adventurous sex, hardier 
(perhaps also harder) emotionally, with more direct interest in 
mechanical (as distinct from human) processes; the female 
professedly gentler, more sympathetic, with more proclivity to 
artistic and humane (or humanitarian) than to mechanical 
pursuits. 

Social Sex Bias. Does each sex tend to like more its own 
sex's occupations? We have classified the occupations into three 
groups as predominantly male, predominantly female, or neutral 
and noted the preference index found in Table 78 for those in 
each group. They are as follows (index of each occupation in 
parentheses) : 

Predominantly male occupations: 

Aviator (9), auto mechanic (16), farmer or rancher (4), factory manager 
(12), fisherman (8), plumber (8), miner (6), explorer (5), auto racer 
(9), building contractor (11), forest ranger (15), dairyman (7), stock 
breeder (11), soldier (8), draftsman (10). These all score male. 
Predominantly female occupations: 
Stenographer (15), teacher (16), nurse (18), dressmaker (25), music 

teacher (13). These all score female. 
Neutral occupations: 
The remaining 30. 

For the predominantly male occupations, the average male 
preference index is 9.2; for the predominantly female the average 
female preference is 17.2. Of the neutral two only are preferred 
by males, "detective" and "architect," the most questionably 



424 SEX AND PERSONALITY 

neutral of the group. The average male preference of these two 
scores is 4.5. The remaining 28 are all female preferences with 
an average female preference score of 9.7. Thus, whereas the 
male tends to confine his preferences to occupations in which he 
already predominates, the female would enter neutral territory 
freely, since she expresses an average preference for 28 neutral 
occupations as great as the average preference of males for even 
dominantly male occupations. 

Further, the expressed female preference for female occupations 
(17.2) is almost twice that of the males for male occupations 
(9.2). That females are particularly enamored of their present 
jobs need not be concluded. The female preference score for the 
female occupations may average higher than the male score for 
male occupations, not because the women love women's jobs, but 
because the men disdain them. Females score high on their own 
jobs because men record no preference for them; men score lower 
on theirs because females are not so chary of occupying male 
preserves. 

Summarizing: both the predictive and the a posteriori analyses 
confirm previous conclusions; females express themselves as more 
inclined than males to indoor, artistic or decorative, and ministra- 
tive occupations; males as more inclined than females to outdoor, 
adventurous, and muscular occupations. 

We find in addition males preferring their own customary 
occupations, females, not only their own but also many male 
occupations. 

How far this indicates protest or interest, and how far the 
preferences in general are genuine or merely professed, the 
statistical contrasts cannot tell us. 

Sections B to Hi Idiosyncrasies, Books, Pastimes, Etc. It 
would be wearisome and unremunerative to discuss the other 
sections of Exercise 5 with the same detail with which we have 
discussed the "sexually" most significant section, A\ especially 
as it needs only a cursory examination of the items to show that 
most of them fall into categories already formulated. It will be 
necessary, therefore, merely to classify them sex-wise, and beyond 
this to consider real or seeming exceptions. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 425 

In classifying items in Sections B to H, both forms, we shall 
exclude those with a preference index of 3 or less (38 out of 257 
items in all). The preferences being comparative, the appear- 
ance of a response as the preference of one sex may mean, not that 
the sex espouses it, but that it exemplifies some quality opposite 
to that favored by the other sex; for instance, unaesthetic objects 
may thus appear as male preferences. In spite of this com- 
plication the sex divergences evidenced from our previous 
exercises clearly obtain. Our classes run as follows: 

A. Distinctive male preferences: 

1. Objects, situations, etc., themselves "male" in character. 

2. Experiences implying adventure or aggressiveness. 

3. Experiences implying an interest in mechanisms and mechanical 

contrivances. 

4. Occupations primarily out-of-doors or involving hard or prolonged 

bodily exercise. 

5. Business, commercial, and political interests. 

6. Experiences the opposite of those favored by females. 

B. Female preferences: 

1. Objects, characteristics, etc., themselves exemplified by females. 

2. Experiences which evoke maternal tenderness or active sympathy. 

3. Aesthetic experiences, including literature. 

4. Domestic experiences. 

5. Social interests. 

6. Experiences the opposite of those favored by the other sex. 

According to whether items of the first six classes do actually 
score as distinctively male, and those of the second six as dis- 
tinctively female, preferences, our hypothesis is supported or 
rebutted. 

Class A 1. Objects, Situations, etc., Primarily Concerning Males. 
In Form A: 

Section B Men with beards. 

Section C Charlie Chaplin. 

Section C Robinson Crusoe, Gulliver's Travels, Adventures of Sherlock 
Holmes, Little Men, Little Lord Fauntleroy, Boy's Life of 
Theodore Roosevelt. 

Sections F, G, and H None. 
In Form B: 

Section B Bald-headed men. 

Section C W. S. Hart, Harold Lloyd. 



426 SEX AND PERSONALITY 

Section Z> -None. 

Section EThe Courtship of Miles Standish, Daniel Boone, Adventures 

of Tom Sawyer, Bob Son of Battle. 
Sections F to H None. 

All these except Little Men, Little Lord Fauntteroy, and The 
Courtship of Miles Standish score male preferences. Of these 
exceptions two are elsewhere accounted for. The third, The Court- 
ship of Miles Standish , has a low female preference index, and it 
will probably be admitted that the title, as well as the contents 
of the poem, have a female equally with a male reference. It is 
scarcely a crucial exception. 

Class A 2. Experiences Associated with Adventure or Aggressive Activity. 

Form A: 

Section B None. 

Section C Detective stories, Adventure stories. 
Section D Hunting, Dare base (?). 
Section E Robinson Crusoe, Gulliver's Travels, Adventures of Sherlock 

Holmes, Boy's Life of Theodore Roosevelt, Rip Van Winkle. 
Section F Tigers, Ships. 
Section G Accidents, Sporting news. 
Section H Hunt lions in Africa, Visit many famous battlefields, See 

how criminals are treated. 
FormB: 

Section B None. 
Section C Sporting pages, Stories of wild animals, W. S. Hart (?), 

Harold Lloyd (?). 
Section D Fishing, Swimming. 
Section EThe Swiss Family Robinson, Daniel Boone, The Call of the 

Wild, Twenty Thousand Leagues under the Sea, Wild 

Animals I Have Known, The Adventures of Tom Sawyer, 

The Spy, Bob Son of Battle, Treasure Island, Two Years 

before the Mast. 

Section F Battle scenes, Lions, Ruins (?). 
Section G Crimes, Fires. 
Section H Climb many lofty mountain peaks. 

All these, except the questionably adventurous Dare base, 
score male preferences. Dare base has a low female preferential 
index of 4. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 427 

Class A 3. Topics Implying an Interest in Mechanics and Mechanical 
Contrivances. 

Form A: 

Section B None. 

Section C Radio magazines, Chemistry. 

Section D Repairing a door latch. 

Sections E, F, and G None. 

Section H Visit many manufacturing plants. 
FormB: 

Section B None. 

Section C Shop work, Popular science books. 

Section D Working with tools, Riding bicycle. 

Section E None. 

Sections F and G None. 

Section H Seeing a ship-building plant and, possibly, Visiting the 
Chicago slaughter-house. 

All these score male preferences. 

Class A 4. Experiences Connected with Outdoor and Vigorous Bodily 
Activities. 

Form A: 

Section B None. 

Section C Adventure stories (?). 

Section D Hunting. 

Section E Robinson Crusoe, Gulliver 9 s Travels. 

Section F Tigers, Ships. 

Section G Sporting news. 

Section H Hunt lions in Africa. 
FormB: 

Section B None. 

Section C Sporting pages, Stories of wild animals. 

Section D Riding bicycle, Fishing, Swimming. 

Section E Swiss Family Robinson, Daniel Boone, The Call of the Wild, 
Twenty Thousand Leagues under the Sea, Wild Animals 
I Have Known, The Adventures of Tom Sawyer, Bob Son 
of Battle, The Spy (?), Treasure Island, Two Years before 
the Mast. 

Section F Battle scenes, Lions, Dogs (?). 

Section G None. 

Section H Climb many lofty mountain peaks. 

All score male preferences. 



428 SEX AND PERSONALITY 

Class A 5. Business, Commercial, Political Interests. 

Form A: 

Sections A to F None. 

Section G Commercial news, News oddities. 

Section H Visit many manufacturing plants (business), Visit many 

battlefields (political). 
FormB: 

Sections B to F None. 

Section G Political news. 

Section H See a ship-building plant, Visit the Chicago slaughter-house. 

All these are male items excepting News oddities, which has a 
low female preferential index, 4. But the item is questionably a 
business, commercial, or political topic. 

Class A 6. Experiences the Opposite of Those Favored by Females. 

There are one or two ugly or disgusting objects or experiences which 
on our hypothesis might be thrown on to the male side by the females' 
recording their dislike for them, e.g., People who use toothpicks (Form B, 
Section B)\ Side-show freaks (Form B, Section J5); Visiting the Chicago 
slaughter-house (Form B, Section H) and possibly Bald-headed men 
(Form B, Section B). These score male preferences. 

We shall now consider the classes of subjects which by hypothe- 
sis should be predominantly female. 

Class B 1. Objects or Topics with a Marked Female Signification. 

Form A: 

Section B Mannish women, Tall women. 
Section E Rebecca of Sunnybrook Farm, Evangeline, Through the 

Looking Glass, etc. 
Form B: 

Section B Thin women, Women cleverer than you. 
Sections C and D None. 

Section E Little Women, Mrs. Wiggs of the Cabbage Patch, Janice 
Meredith, Helen Keller's Story of My Life, Heidi. 

All these have a female preference index. 

Class B 2. Topics Which Evoke Maternal Tenderness or Active Sympathy. 

Form A: 

Section B Babies. 
Section C Movie love scenes. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 429 

Section Z> Pet cats (?). 

Section E Little Men, Little Lord Fauntleroy, Evangeline, Peter Pan and 
Wendy. 

Section F Children. 

Section G None. 

Section U See how criminals are treated. 
FormB: 

Section B School children, Bashful men. 

Section C Love stories. 

Section D Pet canaries (?). 

Section E Water Babies, Helen Keller's Story of My Life, Janice 
Meredith, Mrs. Wiggs of the Cabbage Patch, Heidi, The 
Courtship of Miles Standish, The Prince and the Pauper. 

Section F to G None. 

Section H Study methods of training children. 

All these except "See how criminals are treated" score female 
preferences. But if from schoolchildren the criminal commands, 
not sympathy, but reprobation, the item would not fall into this 
category. The effect of the motives in this class in detaching 
some "ostensibly male" subjects from the male side will be 
noticed; e.g., Little Men, Littte Lord Fauntteroy. 

Class B 3. Aesthetic or Artistic Experiences or Topics Associated with 
Them. 

Form A: 

Section B None. 

Section C Poetry, Dramatics, Ancient languages. 

Section D Collecting flowers, Charades (?). 

Section E (See below). 

Section F Flowers, Clouds. 

Section G Musical events, Theatrical news. 

Section H Spend a day in Westminster Abbey, See London Bridge. 
FormB: 

Section B None. 

Section C Music, Fairy tales, Penmanship (?), Painting. 

Section D Dancing, Gardening (?). 

Section E (See below). 

Section F Sunsets, Trees, Portraits. 

Section G None. 

Section H Study national dress, Visit many picture galleries, Visit 
many cathedrals. 

These all score female preferences. 



430 SEX AND PERSONALITY 

Section E (on book preferences) has been excluded from this 
class, since all good books make some aesthetic appeal; but on the 
other hand, to what extent it is content or style which decides 
the preference it is difficult in any case to say. With most 
schoolchildren we can perhaps safely say that the matter is more 
important than the style. It is noticeable, however, that the 
three topics of a poetic nature, Evangeline, Tales from Shakespeare, 
and Riley's Poems, are female preferences. 

Class B 4. Domestic Experiences or Topics Associated with Them. 

Form A: 

Section 5 Babies. 
Section C Stories of home life. 
Section D Parties and socials, Cooking, Charades, Pet cats, Drop 

the handkerchief (?), Repairing a door latch, Chess (?). 
Section E (See below). 
Section F Flowers, Children, Cats, Fruits (?). 
Section G None. 

Section H See how people prepare their food. 
FormB: 

Section J3 None. 
Section C None (Love stories ?). 
Section D Washing dishes, Pet canaries, Guessing games (?), Ham and 

eggs (?). 

Section J5 (See below). 
Section F and G None. 
Section H Visit the homes of many writers, Learn about Japanese 

houses, Study national dress, Study methods of training 

children. 

With the exception of the rather questionably domestic 
topics, Repairing a door latch and Ham and eggs, 1 these are all 
female preferences. The "mechanical" appeal of the former 
clearly counteracts any domestic appeal it may make. 

As regards Section E (on book preferences) a glance down the 
male and female preference columns in both forms will show that 
the females prefer domestic incident, and the males external 
adventure. 

1 See discussion of Exercise 1, p. 378, also Table 64, p. 376, which show that 
"meat" is more thought about by males than by females. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 431 

Class B 5. Social Experiences and Social Entertainment, Manners, and 
Practices. 

Form A: 

Section C Social problem movies. 

Section D Parties and socials, Hopscotch, Charades, Drop the handker- 
chief, Dare base. 

Section E None. 

Section F None. 

Section G -News oddities (?). 
FormB: 

Section D Jump the rope, Cat and mouse, Guessing games, Picnics. 

Section G Social news. 

Section H Visit the homes of many writers, Learn about Japanese 
houses, Study national dress, Study methods of training 
children. 

These all score female preferences. 

Class B 6. Experiences the Opposite of Those Favored by Males. 
There are none in which this factor seems to be crucial. 

So far, then, these comparisons confirm conclusions as to sex 
differences already suggested from an examination of previous 
exercises. A scrutiny of response scores for 24 items which do not 
fall obviously into these classes reveals no important sex differ- 
ence in addition to, or inconsistent with, those already detected. 

Numerical Excess of Female Preferences. One marked sex 
distinction which seems to pervade this exercise requires com- 
ment. The records show a noticeable numerical excess of female 
over male preferences, and a corresponding excess of male dis- 
likes. In Table 79a, which summarizes the data on this point, the 
term "pronounced like" means that the preference score for 
D and N is of the opposite sex to that for L\ e.g., 4L, +ID, 
+2N means a pronounced female like, while +2L, I/?, 22V 
means a pronounced male like. 

Table 790 shows that in the whole exercise females score 133 and 
males 99 likes, males 135 and females 87 dislikes. Cases of 
" pronounced " likes total 85 for females and 44 for males. The 
reason, however, for the female excess of "likes" does not seem 
to be that there are more things in heaven and earth pleasing 



432 



SEX AND PERSONALITY 



to the female sex, nor merely that the female sex expresses emotion 
more freely; for in that case its dislikes also would be more 
emphatic in contradiction of our records. Taking the sections 
in detail (Table 79i) we get the due. This table shows that the 
differences in numbers of preferences are especially pronounced in 

TABLE 79a. NUMERICAL EXCESS OP FEMALE PREFERENCES 



Preference 


Fo 


rm A 


Fo 


rmB 


T 


otal 




Male 


Female 


Male 


Female 


Male 


Female 


Likes 


48 


68 


51 


65 


99 


133 


Pronounced likes 


22 


40 


22 


45 


44 


85 


Dislikes 


71 


40 


64 


47 


135 


87 


Neutral 


49 


36 


62 


34 


111 


70 

















TABLE 796. NUMBERS OF SEX PREFERENTIAL INDICES IN EXERCISE 5 





Form A 


Form B 


Total 


Section 


Number of P.I.'s 


Number of P.I.'s 


Number of P.I.'s 




Male 


Female 


Male 


Female 


Male 


Female 


A 


9 


16 


8 


17 


17 


33 


B 


5 


7 


4 


8 


9 


15 


C 


6 


8 


7 


6 


13 


14 


D 


6 


14 


9 


11 


15 


25 


E 


10 


13 


12 


11 


22 


24 


F 


3 


5 


5 


3 


8 


8 


G 


3 


3 


3 


3 


6 


6 


H 


5 


6 


3 


8 


8 


14 




47 


72 


51 


67 


98 


129 



some sections, namely, in Sections A and D in Form A, and A, B, 
D y and H in Form B. In the case of Section A (on occupations) 
social sex bias is at work. A tendency of males to prefer male 
occupations, and of females to like male as well as female occu- 
pations, accounts for the female doubling the male preferences 
(33 against 17) . Section D is largely composed of social pastimes 
and occupations of a domestic nature. Section B, with only 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 433 

12 items, we have reason to suppose is in any case sexually the 
least significant section, and Section H (Form B) happens to 
comprise more travel objectives of domestic and aesthetic than of 
business, political, and scientific interest. Had the list included 
such items as "Visit big oil fields" or "Visit the spots where 
dinosaur's eggs were dug up" or "Visit the magnetic pole" or 
"Visit the source of the Nile," we may be fairly certain that 
much of the discrepancy would have vanished. In part a special 
social bias, in part an unbalanced selection, accounts for the 
differences. There remains probably a greater tendency for the 
male to disdain the feminine, than for the female to disdain 
the masculine. 

Male Excess in Neutral Scores. Another feature of the 
exercise as a whole is the excess of male neutral responses as 
indicated in the following figures for Form A and Form B 
combined: 



Categories compared 


Male 


Female 


Number of neutral scores 


111 


70 


Score (sum of plus figures for the male and of minus figures 
for the female) .... 


227 


138 


Average . . . . . 


2. OS 


1.97 









As the neutral response indicates hesitance or indecision, it 
might appear that the male is more reticent in expressing a definite 
preference than the female; or that the female is more decided in 
expressing her likes or dislikes than the male. She is more sure, or 
she says she is more sure, exactly what her feelings are. But 
examination will show that this male neutral tendency is limited 
to certain sections; namely, to Sections A , both forms; 5, Form B ; 
Z>, Form A; , Form B ; and H, both forms. With the exception 
of E (Form B) these are just the sections which have already been 
found exhibiting the most distinctive female preference indices; 
and this has been explained without resort to any general hypothe- 
sis. Reasons have been given why the female should score 
higher like responses in these cases. But an excess of female 
scores of like response automatically involves a corresponding 



434 SEX AND PERSONALITY 

excess of male scores to be shared among dislikes and neutrals. 
In short, the excess of male neutral responses may be occasioned, 
not by any tendency to hesitance or neutrality in general, but 
by the particular circumstances already considered which favor 
a high proportion of female likes. 

Summary. Section A has already been summarized. Analy- 
sis of the multifarious material contained in Sections B to H 
corroborates previous conclusions. We find males expressing 
more taste than females for situations involving males, for adven- 
tures, for outdoor occupations, for machinery, for business, com- 
mercial, and political matters; females expressing more taste 
than males for situations involving females, for indoor and social 
interests, and for experiences evoking active sympathy or 
maternal tenderness. 

A female tendency to express likes more emphatically and 
frequently than males, and a male tendency to more dislikes and 
more neutral attitudes than females, largely explicable by unbal- 
anced selection of items and by acquired social bias, is consonant 
with a distinctive female proneness to express enjoyment more 
freely. 

EXERCISE 6: PERSONAGES AND OPINIONS 

This exercise has two sections, one calling for expression of 
attitude toward selected historical personages, the other for judg- 
ment of truth or falsity of certain statements. Responses to both 
sections were analyzed for the following subjects: 121 boys and 
122 girls in the seventh grade, 102 boys and 118 girls in high 
school, and SO men and SO women in Stanford University; a total 
of 563. 

Personages. Table 80 gives the preference indices of the 
55 items contained in the two forms of the test. 

The following points emerge: 

1. No woman's name has a male index. 

2. One woman (Cleopatra) has a neutral score. The sex-to-sex 
tendency is here, it seems, counterbalanced by the tendency of 
females remarked elsewhere to be more severe than males on 
"loose" living. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 435 



TABLE 80. PREFERENCE INDICES OF HISTORICAL PERSONAGES 

Form A 



No. 



Character 



1 JackDempsey +10 

2 Kit Carson + 7 

3 P. T. Barnum + 6 

4 Robert G. Ingersoll + 5 

5 Ulysses S. Grant + 4 

6 Bismarck 4-3 

7 Daniel Boone + 3 

8 Oliver Cromwell . -f- 2 

9 Mussolini -f 2 

10 Billy Sunday -f 2 

1 1 Judge Ben Lindsay +1 

12 Theodore Roosevelt -f 1 

Female 

preference 

index 

1 Florence Nightingale . 7 

2 Jane Addams . 4 

3 Hearst, the publisher 3 

4 Aaron Burr 3 

5 Woodrow Wilson . - 3 

6 Christopher Columbus 2 

7 Aimee McPherson. ... 

8 Lenin 

9 Thomas Jefferson 

10 Jefferson Davis. . . 

1 1 Congressman Volstead . 

Neutral index 

Cleopatra .. 

Wellington .. 

Booker T. Washington. 

Herbert Hoover. .. . ... 

Lloyd George. . 



Male 

preference 
index 



Form B 



No. 
2 

6 

8 

9 

10 

11 

12 



9 
10 
11 
12 
13 



Character 



Male 

preference 
index 



"Babe" Ruth. 
John L. Sullivan 
*'" Cobb 



Hannibal 

John Brown ......... 

Caesar ............ 

Marconi ............ 

Tom Paine ......... 

Charles Darwin 
Anthony Comstock. 
Napoleon ... . . . 

Brigham Young ..... 



Cornwallis 

John Alden .... 
Queen Victoria. 
Galileo. 



Carrie Chapman Catt 

Eugene Debs 

Mohammed , 

Daniel Webster 

Kaiser Wilhelm 

Mary Baker Eddy 

Henry Ford 

Robert E. Lee 

Abraham Lincoln . . . 

President Coolidge.. 
Jesse Tames 



4-10 
+ 10 

I 

Female 

preference 

index 

- 6 

- 4 

- 3 

- 2 

- 2 

- 2 

- 1 

- 1 

- 1 
Neutral index 






436 SEX AND PERSONALITY 

3. There are eight generals. Of these only two and those 
were generals who suffered defeat have a female index; namely, 
Cornwallis and Lee. Cornwallis, perhaps the least successful 
of the eight, has the highest female preference score. This 
supports the view that females are more moved by active sym- 
pathy with misfortune, while aggression and enterprise appeal 
more to males. 

4. Unfortunates other than generals are also female prefer- 
ences: Columbus, Wilson, Galileo, Debs, and Kaiser Wilhelm, all 
of whom suffered either persecution, misunderstanding, or a 
reversal of fortune. Possibly we should include Aaron Burr in 
this dass. This again suggests a distinctive female sympathy. 
Hannibal and John Brown are the nearest approach to exceptions 
on the male index list; but perhaps the military exploits of 
Hannibal, and the sporting end of John Brown, suffice to win for 
them distinctive male preference. 

5. All the notorious pugilists and sports champions (and one 
hunter) score very pronounced male preferences: Dempsey, 
"Babe" Ruth, Kit Carson, and "Ty" Cobb. 

6. Of the women characters (none of whom are on the male 
side) the most well-known philanthropists Florence Nightingale 
and Jane Addams win the highest, and those noted for more 
questionable qualities Mary Baker Eddy and Aimee McPherson 
win lower, female preference scores. The active sympathizers 
with human suffering appeal most distinctively to the female sex. 

7. Individuals whose characteristic feature is defiance of con- 
ventions score male: Tom Paine, Ingersoll, and Judge Lindsay. 

These preference differences confirm previous conclusions. 
Wellington and Mohammed are the only surprises. 

An excess of female omissions to respond in 43 out of 55 cases 
the proportion was higher for females may be due to the fact 
that fewer girls interest themselves in famous male personages. 
(There are only seven women characters in the list.) 

Opinions. The statements presented in this section to be 
marked T (true) or F (false), may, with the exception of two or 
three, be considered matters of opinion rather than of knowledge, 
and the section becomes a study of sex prejudices. That a sex 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 437 

tends to favor its own interests explains the response scores in 
the following items: 

Form A: 

Girls are naturally more innocent than boys. T, 3; F, +3. 
There should be perfect equality between men and women in all things. 
T, 3; F, +3. (Female protest against existing sex discrimination.) 
FormB: 
Married women ought not to be permitted to teach school. T, +1; 

F, -1. 

Women are purer than men by nature. T, 2; F, +2. 
Wealth, honor, and power usually go to those who deserve them. 
T, +2;F, -2. 

Other items testify to the gentler and more sympathetic tend- 
encies of the female, or the harder and more aggressive attitudes 
of the male: 

Form A: 

Children should be taught never to fight. T, -3; F, +3. 

We should never give to beggars. T, +3; F, 4. 
FormB: 

An ugly face usually goes with a kind heart. T, 2; F, +3. (We do 
not know how many of the subjects were ugly, but an ugly face is 
a more serious matter with a woman than with a man, and she sym- 
pathizes more with the troubles of its owner.) 

The hanging of murderers is justifiable. T, +4; F, 3. 

Hunting and fishing are wrong because cruel. T, 3; F, +3. 

The United States should adopt a more aggressive foreign policy. 
T, +3;F,0. 

Of the sex responses in the remaining items, none seems to oppose, 
some to support, our previous conclusions. 

Summary. (1) Among well-known historical and contem- 
porary personages, females show a distinctive preference for 
women, unfortunate people, and philanthropists; males for 
successful generals, sports heroes, and defiers of convention. (2) 
The section on opinions indicates that females tend to favor 
opinions which favor them, and that they subscribe to kindly, 
males to rougher, sentiments. 



438 SEX AND PERSONALITY 

EXERCISE 7: INTROVERTIVE RESPONSE 

The exercise contains 42 questions in each form, with "Yes" 
or "No" answers, on tastes, habits, emotional and imaginative 
tendencies, facts of experience, and a few other topics, many of 
the answers demanding self -scrutiny. In general they are of the 
type commonly used in tests of introversion or neurotic tendency. 
The subjects were the same as for Exercise 6, except that the num- 
ber of high-school girls was 102 instead of 118. 

We may arrive at an index of preference (for an affirmative 
answer) by adding the figures under "Yes" and "No" and 
attaching to the sum the sign under "Yes," since the weights for 
"Yes" and "No" responses are nearly always the same and 
never differ by more than 1 ; e.g. : 

"Do people ever say you talk too 

much?" +2 Yes -INo gives a P. I. +3 (male) 

"Are you happy most of the time?" 3 Yes +3 No gives a P.I. -6 (female) 

Table 81 gives the items of each form arranged in order of 
preference index. The latter is starred whenever the "Yes" 
response indicates introvertive or neurotic tendency; if the item is 
doubtful in this respect the index is followed by a question mark. 

1. A male distinctive tendency toward the aggressive, the 
independent, the adventurous, accounts for male outnumbering 
female affirmative answers to some 12 questions: Form A, 15, 
17, 30, and 35; Form B, 1, 4, 11, 24, 25, 28, 29, and 33. 

2. The female distinctive tendency to feel or affect more fear, or 
at least to admit its influence on behavior, accounts for female 
outnumbering male affirmative answers to Questions 2 and 13 in 
Form A, and Questions 2, 6, 12, and 22 in Form B. 

3. The distinctive female care for personal appearance explains 
the female affirmatives in Form A, Questions 1,3, and 6. 

4. A keener female active sympathy occasions the excess of 
female affirmative answers to Questions 13 and 18 in Form B. 

5. The female preponderance in Form A, Question 4, testifies 
again to the distinctive female penchant for social gatherings. 

All these differences accord with principles by this time 
familiar. A question of interest is whether the remaining 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 439 
TABLE 81. PREFERENCE INDICES OF INTROVERTIVE RESPONSES 



Preference index 



M-P test items 



Male Female 



Form A. 

1. Are you extremely careful about your manner of dress? 12* 

2. Are you often fnghtened in middle of night? 10* 

3. Do you always remember to brush your teeth? 8? 

4. Do you like to go to parties, dances, or other social affairs? ... . 7 

5. Is it easy for you to get up as soon as you wake?. 8 

6. Would you like to wear expensive clothes? . . 6* 

7. Did you ever have imaginary companions? 6* 

8. Do people often say you are too noisy? 6* 

9. Do you rather dislike to take your bath? 6? 

10. Are you much embarrassed when you make a grammatical mistake? .. 6* 

11. Have you ever kept a diary? . . 6* 

12. Do you ever feel that you are about to "go to pieces"? 6* 

13. Are you often afraid of the dark? .. .. 6* 

14. Can you usually sit still without fidgeting? . . 6 

15. As a child were you extremely disobedient? 6* 

16. Do people nearly always treat you right? . 6 

17. Were you ever expelled from school, or nearly expelled?. 6* 

18. Have you often been punished unjustly? 6* 

19. Do you often get cross over little things? . . 6* 

20. Can you stand as much pain as others can? 6 

21. Have you found school a hard place to get along in? 6* 

22. Do you nearly always prefer for someone else to take the lead? 5* 

23. Have you often fainted away? 4* 

24. Do you usually enjoy your meals? . . 4 

25. Have you the habit of biting your finger nails? 4* 

26. Does it make you angry for people to hurry you? 4* 

27. Do you feel tired a good deal of the time? 4* 

28. Do you hear easily when spoken to? . 4 

29. Do people ever say that you talk too much? . 3 

30. Have you been bossed too much for your own good? ... . 2* 

31. Do you feel yourself to be lacking in self-control? . 2* 

32. Do you shrink from facing a crisis or difficulty? 2* 

33. Do you worry much over possible misfortunes? 2* 

34. Are you worried when you have an unfinished job on your hands? 2* 

35. Do you ever dream of robbers? 2* 

36. Do you ever have the same dream over and over? 2* 

37. Do you ever walk in your sleep? 2* 

38. Can you do good work while people are looking at you? 2 

39. Do you feel like jumping off when on a high place?. . .2* 

40. Do you like most people you know? . . 1 

41. Do you work mostly by fits and starts? . 

42. Are you careful of your personal belongings? 



440 SEX AND PERSONALITY 

TABLE 81. PREFERENCE INDICES OF INTROVERTIVE RESPONSES. (Continued) 



M-F test items 



Form 

1. 

2. 

3. 

4. 

5. 

6. 

7. 

8. 

9. 
10. 
11. 

12. 
13. 
14. 
IS. 
16. 
17. 
18. 
19. 
20. 
21. 
22. 
23. 
24. 
25. 
26. 
27. 
28. 
29. 
30. 
31. 
32. 
33. 
34. 

35. 
36. 
37. 
38. 
39. 

40. 
41. 
42. 



B 

Can you stand the sight of blood? 

Do you have a great fear of fire? 

Are your feelings often very badly hurt? 

Were you ever fond of playing with snakes? 

Do you know anybody who is trying to harm you? 

Does it make you uneasy to cross a wide street or open square? . . . 
Do you often have a hard time making up your mind about things? 

Do you think you are getting a square deal in life? 

Are you happy most of the time? 

Do you usually get to do the things that please you most ? . ... 
Do you say what you consider the truth regardless of how others 

may take it? 

Have you been troubled by a fear of being crushed in a crowd? . 
Does the thought of hurting a person or animal give you great pain? 

Do you prefer to go without breakfast? 

Do you always get on well with others? 

Have others of your age usually let you play with them? ... 
Do you think people like you as much as they do others?. . 

Do you like people to tell you their troubles? . . 

Do you dislike the company of the opposite sex? 

Do people ever say you are a bad loser? 

Have you often been bothered by blushing? 

Do you dread to have your picture taken? 

Do you think a thing over carefully before you do it? 

Did you ever run away from home? 

Do you like to tease people till they cry? 

Do you usually feel well and strong? 

Do things ever seem to get misty before your eyes? 

Do people you are with say you quarrel too much? 

Do people find fault with you too much? 

Do you like to be the center of attention in a crowd? 

Do you often introspect or analyze your feelings? 

Do you feel bored a large share of the time? 

Do you prefer to work a thing out for yourself rather than ask help? 
Are you often bothered by the feeling that people are reading your 

thoughts? 

Do you get tired of people easily? 

Would you rather be alone than with someone? 

Do you like to be praised and made much of? 

Do you sometimes wish you had never been born? 

Do you ever imagine stories to yourself so that you forget where you 

are? 



Do you prefer to be with older people? 

Do you like to tell your troubles to others? 

Do you take life so seriously that it sometimes bothers you?. 



Preference index 



Male 



12 



10? 
8* 



4* 
4* 



4* 
4* 

4? 



4* 
4* 



2* 
2* 
2* 



2* 
2* 



2* 



Female 



12* 
12* 



8* 

8* 

8 

6 

6 



6* 

6* 

6? 

6 

6 

6 

4? 



4* 
4* 



4 

4* 



2* 



2* 
2* 

2* 

0? 
0* 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 441 

responses which do not fall clearly into any of these pigeon- 
holes convey evidence of any fresh sex characteristics not yet 
encountered. 

A number of answers seem to fit the view that the male 
presumably the schoolboy finds, or says that he finds, more 
difficulties in direct association with his fellows than the girl does. 
Consider the following items: 

Have you found school a hard place to get along in? Male index, 6. 

(Index of male preference for an affirmative answer.) 
Do you know anybody who is trying to harm you? Male index, 8. 
Do you always get on well with others? Female index, 6. 
Do you feel that you are getting a square deal in life? Female index, 8. 
Are you happy most of the time? Female index, 6. 
Do you usually get to do the things that please you most? Female 

index, 6. 

Have others of your age usually let you play with them? Female index, 6. 
Do you think people like you as much as they do others ? Female index, 6. 
Do people nearly always treat you right? Female index, 6. 
Do people you are with say you quarrel too much? Male index, 4. 
Would you rather be alone than with someone? Male index, 2. 
Do you prefer to be with older people? Male index, 2. 

These answers seem to suggest that the young male finds mixing 
happily with his fellows a more difficult and troublesome affair 
than the schoolgirl does. Is it another effect of the greater 
roughness and aggressiveness of boy society, and the gentler 
and more sympathetic attitude of girls? The only two responses 
that at first sight may seem exceptions are: 

Do you sometimes wish that you had never been born? Female index, 2. 
Do people often say you are too noisy? Female index, 6. 

But of these the first has a very small female preference score, and 
the second may merely result from the girl's being expected to be 
quieter than the boy, and being the more reproved for noisiness. 
Of the remaining items, in some 13 the sex preferences are 
explicable on the assumption that girls find it easier than boys to 
admit emotional and physical frailties. The following 17 ques- 
tions concern such weaknesses admitted in only four cases by a 
higher proportion of males than of females: 



442 



SEX AND PERSONALITY 



M-F test items 



Female 
P.I. 



Male 
P.I. 



Form A: 
Are you much embarrassed when you make a grammatical 

mistake? 6 

Do you ever feel that you are about to "go to pieces "? .... 6 

Do you often get cross over little things? 6 

Can you stand as much pain as others can? . . 6 

Have you often fainted away? . . 4 

Does it make you angry for people to hurry you? 4 

Do you feel tired a good deal of the time? 4 

Do you feel yourself to be lacking in self-control? 2 

Do you worry much over possible misfortune? 2 

Are you worried when you have an unfinished job on your 

hands? 2 

Do you feel like jumping off when you are on a high place? 2 
Form B: 

Are your feelings often badly hurt? 12 

Do you often have a hard time making up your mind about 

things? 8 

Have you often been bothered by blushing? 4 

Do you usually feel well and strong? 4 

Do things ever seem to get misty before your eyes? 4 

Are you often bothered by the feeling that people are reading 

your thoughts? 2 



Again, more instances of abnormal or "queer" behavior are 
admitted by females: e.g., imaginary companions, dream recur- 
rences, sleep walking, the impulse to jump off high places, and 
absorption in daydreaming. Now, if these responses were true, 
we should certainly expect the female sex to prove more intro- 
spective than the male, for most of them accord with the intro- 
spective temperament. But when we come to the direct question , 
"Do you often introspect or analyze your feelings?" more males 
answer in the affirmative. Introspection conveys no stigma of 
weakness; and there is no reason why the male should not admit 
it. 

In the two forms there are some 57 items for which a "Yes" 
answer would class as the more introvertive response in the 
commonly accepted sense of that term. Of these, 35 have a 
female, 22 a male, preference index. The difference is probably 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 443 

significant and is in accord with the usual finding that females 
score somewhat more introverted than males on tests like those 
of Laird, Thurstone, and Bernreuter. 

Summary. To these questions affirmative answers which 
suggest a spirit of adventure or independence in the subject are 
predominantly male; admissions of fear and of humanitarian 
tendencies, of care for personal appearance, or of liking for social 
gatherings, are predominantly female. Several answers suggest 
that boys find school life harder than girls do, perhaps because 
of the greater roughness of boys at school. On the whole, females 
admit more emotional and slightly more physical frailties than 
males and more abnormalities of emotion and "mental" experi- 
ence. They are more given to introvertive response than the 
males. 

SUMMARY AND CONCLUSIONS 

Before concluding this inquiry with a brief discussion of the 
outcome it is desirable, even at the risk of seeming tiresomely 
repetitive, to recapitulate its main findings, exercise by exercise; 
for their strength rests upon cumulative evidence rather than 
upon any single decisive experiment. Taken serially they run 
as follows: 

Exercise 1 : Word Association. 

1. If we take the associated words absolutely, females pick 
more often than males upon terms for domestic things or things 
suggestive of kindly or sympathetic activities; and much more 
often upon terms for articles or qualities of adornment, and for 
colors; males more often upon scientific and business terms and 
particularly upon terms suggestive of excitement and adventure, 
and upon words for foods; and each sex "prefers" names of per- 
sons of its own sex. 

2. If we take the items which show the most sex-contrasted 
responses, divergences follow the same lines and also indicate a 
male preference for words signifying machinery, common tools, 
and outdoor pursuits. 



444 SEX AND PERSONALITY 

Exercise 2 : Ink-blot Association. 

3. Response words most obviously associated with machinery 
or science and with outdoor activities and adventure are picked 
upon more often by males; more of those connected with domestic 
occupations and with aesthetic experience or personal adornment 
by females. 

4. When we consider the items which show the most sex- 
contrasted responses, the term which the stimulus figure evokes 
is in almost every case one connected with a common occupation 
distinctive of the responding sex. 

Exercise 3 : Information. 

5. Females are more correctly informed about domestic 
occupations, domestic and individual embellishment, etiquette, 
fictional literature, certain points of musical technique, color 
shades and differences, and on topics that appeal to active 
sympathy and maternal interest. Males are more correctly 
informed about political, business, economic, scientific, and 
physical facts; about exploits, adventures, and inventions; and 
on topics that evoke aggressive and active bodily propensities. 

6. This divergence between the sexes is most marked in 
knowledge of facts or events that interest aggressive or adven- 
turous dispositions on the one hand, and maternally tender or 
sympathetic dispositions on the other. 

Exercise 4 : Emotional and Ethical Response. 

7. Females express the most distinctive degree of anger on 
occasions of very unsympathetic or cruel treatment of human 
beings where help or sympathy is meet. 

8. Females tend to show a more distinctive degree of anger 
over school offenses and "social" vexations than over business 
and extradomestic troubles. 

9. The sexes diverge more in the fear than in the anger 
responses. 

10. Within the set of fear-inspiring objects and situations 
presented, the more fearsome the object the greater the female 
distinctive fear. 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 445 

11. Females express more disgust than males in general, but 
particularly at "disgusting" male practices (repugnant habits of 
dress and person), at coarse language, and at sexual immorality. 

12. Females express in general more pity than males, but most 
for the weak, helpless, and visibly distressed, especially where 
the object is a human being or a creature attractive in appear- 
ance, and for cases of female more than of male distress. 

13. Females are in general more condemnatory than males, but 
more noticeably in petty offenses and in offenses more common in 
males. In negligible offenses on the one hand and very serious 
offenses on the other the sex distinction is inconsiderable. 

14. The severity of male censure increases more than that of 
female censure with increased gravity of the offense. 

15. Of the alternative occupations and objects presented for 
choice in the last section of the exercise, males distinctively prefer 
the out-of-door and adventurous, and the useful rather than the 
decorative; females prefer indoor and urban conditions of living, 
and things attractive in appearance rather than the useful. 

16. Females express more liking for working with men than 
men for working with women. 

17. Females tend to express a higher degree of the four emotions 
(particularly of disgust) and of moral censure, than males express; 
but proportionately less in moral censure and anger than in 
pity, fear, and disgust. 

18. In general, males record more defect of emotion and of. 
moral censure than females, particularly of disgust and fear, with 
pity, moral censure, and anger following in that order. 

19. The interval between female excess and male defect aver- 
ages greatest with disgust, least with anger and moral censure. 

Exercise 6 : Interests. 

20. The occupations for which females express a distinctive 
preference are the indoor, artistic and decorative, and the 
directly "ministrative"; distinctively preferred by males are 
those entailing adventure, bodily risk, and muscular strength or 
prolonged exertion, 



446 SEX AND PERSONALITY 

21. Males distinctively prefer occupations undertaken pre- 
dominantly by males; females like male as well as female 
occupations. 

22. Females record a higher distinctive preference for pre- 
dominantly female occupations than males for predominantly 
male occupations; but also, on the whole, as high a preference for 
mixed occupations as males for male occupations. 

23. Throughout Sections B to H of this exercise a sex tends to 
prefer objects or topics which more particularly engage or con- 
cern members of that sex. 

24. Among the various types of material presented for response 
in Sections B to H , males are found to express distinctive prefer- 
ence for experience implying adventure, aggressiveness, or interest 
in mechanical contrivances; for out-of-door and physically strenu- 
ous occupations; and for business, commercial, and political 
interests. Females express preference for experiences evoking 
maternal tenderness or active sympathy, and for aesthetic and 
domestically social experiences. 

25. In the aggregate, female likes considerably exceed male 
likes, and male dislikes exceed female dislikes, but this holds of 
some sections more than of others. In three sections of one form 
male exceed female preferences. 

26. In the aggregate, male neutral scores (neither like nor 
dislike) exceed female considerably; but (with one exception) this 
excess is confined to those sections in which female likes are most 
distinctive. 

Exercise 6 : Personalities and Opinions. 

27. Females show a distinctive preference for women, unfor- 
tunate people, and philanthropists; males for successful generals, 
sports heroes, and defiers of convention. 

28. Females distinctively believe statements favorable, and 
disbelieve statements unfavorable, to the female sex; they sub- 
scribe to kindly, males to rougher, sentiments. 

Exercise 7 : Introvertive Response. 

29. Males affirm distinctively tastes and habits that involve 
adventure or courage; females habits or experiences that imply 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 447 

timidity, active sympathy, and care for personal appearance; and 
with one or two exceptions they more readily confess weaknesses 
in emotional control and (less noticeably) of physique. Females 
also admit more psychic abnormalities. 

30. Females more often than males give the introvertive type 
of response. 

We may now consider two questions to which the present 
findings give rise: (1) Can we extract from them a single prime 
principle of sex difference at once not too vague to be ambiguous, 
and not so particular as to be insignificant? (2) What, so far as 
our evidence goes, appears to be the relation of the differences we 
have enumerated to nature and nurture, to endowment and 
environment? We shall take these questions in succession. 

1. Is there one dominant principle? 

It is obvious that from whatever point we have started, whether 
from the knowledge shown by the sexes or from their associations 
or their likes and dislikes for people, vocations, pastimes, books, 
or objects of travel; or whether we have explored directly or 
deviously their emotions, tastes, opinions, and inner experiences, 
we have found ourselves arriving at much the same conclusions 
all our ways have led to Rome. But the final scene has two 
aspects two sides of the same picture one showing differences 
in the direction of interest, the other differences in the direction 
of emotions and impulses. 

From whatever angle we have examined them the males 
included in the standardization groups evinced a distinctive inter- 
est in exploit and adventure, in outdoor and physically strenuous 
occupations, in machinery and tools, in science, physical phe- 
nomena, and inventions; and, from rather occasional evidence, in 
business and commerce. On the other hand, the females of our 
groups have evinced a distinctive interest in domestic affairs and 
in aesthetic objects and occupations; they have distinctively 
preferred more sedentary and indoor occupations, and occupations 
more directly ministrative, particularly to the young, the help- 
less, the distressed. Supporting and supplementing these are 
the more subjective differences those in emotional disposition 
and direction. The males directly or indirectly manifest the 



448 SEX AND PERSONALITY 

greater self-assertion and aggressiveness; they express more 
hardihood and fearlessness, and more roughness of manners, 
language, and sentiments. The females express themselves as 
more compassionate and sympathetic, more timid, more fas- 
tidious and aesthetically sensitive, more emotional hi general 
(or at least more expressive of the four emotions considered), 
severer moralists, yet admit in themselves more weaknesses hi 
emotional control and (less noticeably) in physique. 

But we must define some of our terms more precisely, for 
instance, "aggressiveness" and "self-assertion." The evidence 
is for initiative, enterprise, vigorous activity, outdoor adventure; 
"aggressiveness" need not imply selfishness or tyranny or unfair 
attack. The compassion and sympathy of the female, again, 
appears from the evidence personal rather than abstract, less a 
principled humanitarianism than an active sympathy for palpable 
misfortune or distress. In disgust, in aesthetic judgment, and 
hi moral censure, the evidence is rather for the influence of fashion 
and of feeling than of principle or reason. Our evidence need not 
imply the possession of a "truer" taste or a more discerning 
conscience. 

But in asking how deep these sex distinctions go we reach our 
second question: What appears to be the relation of our main sex 
difference to nature and nurture, to endowment and environment? 

The question is not, let us remind ourselves, whether this or 
that trait is innate or acquired, for every human act or thought is 
both, but whether the actual sex differences we are discovering 
are ascribable to biological (genetic) factors dividing the sexes 
or to sex differences in their training and environment. So far 
as the evidence of our experiment goes, we are not justified in 
ascribing the manifest differences to one alternative exclusively. 
Certainly we do not have enough evidence to exclude the gross 
physiological differences between the sexes from any part in 
determining the distinctive preference of the male for heavy 
muscular work and of the female for less active occupations, or in 
determining her greater sympathy for the young and weak or 
her greater interest in home life, with the relegation of outside 
interests to the male. To actual or anticipated childbearing and 



SEX TEMPERAMENTS AS REVEALED BY THE M-F TEST 449 

motherhood differences physiologically determined we have 
found no reason to deny a part in determining differences in overt 
habits and emotional dispositions. And in the present state of 
our ignorance it would be even more rash to deny the possible 
influence upon sex temperaments of the manifold differences 
between the sexes in their endocrine equipment and functioning. 
Whatever our view as to the innateness of the distinctive 
tendencies, at least as to maternal tenderness in the one sex and 
comparative aggressiveness in the other, our experimental 
evidence is inconclusive. However, when we examine the more 
direct manifestations of these and other contrasting tendencies 
in our exercises, and consider how any particular manifestation 
comes about, the power and reach of what we have named cul- 
tural sex bias, its many plain and subtle effects on the upbringing 
and environment of the sexes within the groups we are con- 
sidering, keep coming to one's mind. In so many ways too famil- 
iar to realize, each sex gives and receives such different treatment 
as largely to explain the divergences in expression or in fact 
revealed by the material we have studied. Singularly powerful 
in shaping our development are other people's expectations of us, 
past and present, as shown by their practice and their precept. 
Whether the boy is innately more aggressive and fearless, more 
handy with the electric lighting than with the cooking stove, 
more interested and informed about public affairs and about 
science, more active and enterprising physically; and whether the 
girl is by nature more sympathetic, gentle, timid, fastidious, more 
attracted to pots and pans than to rods and guns, more punc- 
tilious in dress, personal appearance, manners, and language; at 
any rate society in the shape of parents, teachers, and one's own 
fellows of whichever sex expects these differences between the 
sexes, and literature reflects them. Irresistibly each sex plays 
the role assigned, even in spite of its own protests. The con- 
sequence is that throughout these several exercises, however 
statistically consistent the distinctive sex responses may prove, we 
cannot tell how deep the difference lies or how the deeper and 
shallower factors combine. And here we must be content to 
leave the problem, for it is clear that the deciding answer can be 



450 SEX AND PERSONALITY 

wrested, not by a more meticulous struggle with this one set of 
exercises administered to groups comparatively homogeneous, but 
from: (1) parallel examinations of socially and racially different 
groups widely different in social tradition and circumstance, and 
(2) combined psychological and biological case studies of extreme 
deviants in sex temperaments within a given culture. 



CHAPTER XVII 

INTERPRETATIONS AND CONCLUSIONS 

THE SIGNIFICANCE or M-F DIFFERENCES FOR PERSONALITY 

Masculinity and femininity are important aspects of human 
personality. They are not to be thought of as lending to it 
merely a superficial coloring and flavor; rather they are one of a 
small number of cores around which the structure of personality 
gradually takes shape. The masculine-feminine contrast is 
probably as deeply grounded, whether by nature or by nurture, as 
any other which human temperament presents. Certainly it is 
more specifically rooted in a structural dichotomy than the 
cycloid-schizoid or the extrovertive-introvertive contrasts. 
Whether it is less or more grounded in general physiological and 
biochemical factors than these remains to be seen. In how far 
the lines of cleavage it represents are inevitable is unknown, 
but the possibility of eliminating it from human nature is at least 
conceivable. The fact remains that the M-F dichotomy, in 
various patterns, has existed throughout history and is still 
firmly established in our mores. In a considerable fraction of the 
population it is the source of many acute difficulties in the indi- 
vidual's social and sexual adjustment and in a greater fraction it 
affords a most important impetus to creative work and happiness. 
The indications are that the present situation, together with the 
problems it raises for education, psychology, and social legislation, 
will remain with us for a long time to come. 

As long as the child is faced by two relatively distinct patterns 
of personality, each attracting him by its unique features, and is 
yet required by social pressures to accept the one and reject the 
other, a healthy integration of personality may often be difficult 
to achieve. Cross-parent fixations will continue to foster sexual 
inversion; the less aggressively inclined males will be driven to 

451 



452 SEX AND PERSONALITY 

absurd compensations to mask their femininity; the more 
aggressive and independent females will be at a disadvantage 
in the marriage market; competition between the sexes will be 
rife in industry, in politics, and in the home as it is today. 

Even if it could be shown that the malleability of personality 
is such as to make the adoption of a single ideal pattern of tem- 
perament feasible, no one knows whether the consequences would 
be more desirable than undesirable. So far only one single- 
standard society has been described for us, that an extremely 
primitive one consisting of but a few hundred individuals living 
in the wilds of New Guinea. Mead's description 1 of this society, 
challenging as it is, offers no very convincing evidence that a 
system of unipolarity reduces the difficulties of individual adjust- 
ment. Conceivably, in a more complex society it might increase 
them. It is possible that in an enlightened culture, no longer 
held in leash by traditions and taboos, dual patterns of sexual 
temperament are an aid in the development of heterosexuality. 

But it is not our purpose to defend the prevailing ideals with 
respect to sex temperaments. The irrelevance and absurdity of 
many of their features are evident enough. That in most 
cultures they have been shaped to the advantage of the physically 
stronger sex is obvious. It does not necessarily follow that a 
dichotomy of temperaments is per se an evil to be got rid of. In 
any case it is not the business of the scientist either to condemn 
or to praise any given type of human behavior. His task is to 
understand it. The application of his findings to social better- 
ment he is willing to leave to the social reformer, but with respect 
to the personality problems with which we are here concerned, he 
knows that intelligent reform will have to await the establishment 
of a substantial body of knowledge which does not now exist. 

THE NEED FOR MORE ADEQUATE DESCRIPTION OF SEX 
TEMPERAMENTS 

The first step in the investigation of the sex temperaments is to 
make possible their more adequate description and more exact 
identification. We have shown that descriptions based upon 

1 MEAD, MARGARET, Sex and temperament in three primitive societies , 335 pp., 
Morrow, 1935. 



INTERPRETATIONS AND CONCLUSIONS 453 

common observation are often contradictory and that even a 
subject's intimate friends register little agreement in rating him 
for degree of masculinity and femininity. This state of affairs 
betokens the vagueness of current ideas with respect to what 
constitutes the masculine or feminine temperament and the chaos 
of opinion with regard to what is valid evidence of its existence. 
Three sources of confusion may be briefly mentioned. 

1. Erroneous ratings may result from the too ready acceptance 
of overt behavior as the criterion. In this respect the investi- 
gator of personality or character is at a disadvantage in com- 
parison with the investigator of intelligence or other abilities. 
Subjects do not often try to hide their intelligence and they are 
unable to hide very effectively their stupidities, but character 
and personality can be rather successfully simulated. Within 
limits the dishonest can simulate honesty, hatred can be hidden 
under honeyed words, anger can be disguised, the introvert can 
force himself to behave as an extrovert, the homosexual may 
deport himself so normally as to remain undetected in our midst. 

2. Errors may be due to lack of a sufficiently large sampling of 
observational data. The teacher's contacts with her pupils are 
limited to certain types of situations. The same is true of our 
contacts with most of the people we know. 

3. Among the hardest errors to eliminate are those that arise 
from traditional biases, such as the notion that the masculine 
temperament nearly always goes with a particular type of voice, 
physique, carriage, manner of dress, or occupation. There are 
doubtless other biases more or less peculiar to the individual or 
to the class to which he belongs, varying according to whether 
he is male or female, masculine or feminine, young or old, strongly 
or weakly sexed, etc. 

It is evident that no clear delineation of sexual temperaments is 
possible on the basis of uncontrolled observation. The M-F test 
is an attempt to remedy this situation. Its scientific intent is to 
free the concepts of masculinity and femininity from the irrele- 
vancies and confusions which have become attached to them as 
the result of superficial consideration of everyday behavior. It is 
necessary to go back of behavior to the individual's attitudes, 



454 SEX AND PERSONALITY 

interests, information, and thought trends, which constitute the 
real personality behind the front presented to his fellows. 

That the purpose of the test has been accomplished only in part 
hardly needs to be said. Our sampling of the universe of mental 
attitudes and interests which differentiate the sexes is far from 
adequate. The sampling used has not been validated by item 
counts for sufficiently large populations. Numerous questions 
remain unanswered with respect to the selection of test items, the 
best method of weighting responses, and the most meaningful 
kinds of scores to employ. The defects of our technique will be 
remedied by experiment, the technique itself seems to us inescap- 
able however much it may require supplementation by direct 
experimental procedures. 

M-F RATINGS OF POPULATION GROUPS BY PSYCHOLOGISTS 

The difficulties encountered by one who is called upon to rate 
persons for masculinity and femininity are well illustrated by an 
experiment we have made in which psychologists were asked to 
rank certain populations in regard to this aspect of personality. 
The rankings were made by 21 male and 21 female psychologists, 
selected on the basis of their researches in the field of personality. 1 
To each were sent four lists of populations with the request 
that the populations of each list be ranked in order according 
to estimated mean with respect to "mental masculinity/' the 
most masculine population to be rated 1 . We purposely refrained 
from defining the term "mental masculinity" beyond saying 
that we had in mind primarily "masculinity in interests, attitudes, 
and thought trends." (See Table 82 for populations rated.) 

Let us examine first the reliability of the ratings. The averages 
of the intercorrelations of the individual raters are as shown at 
the top of page 455. The highest agreement is among the 
male judges of male occupational groups; the next highest is 
for female judges of both male and female occupational groups. 

1 We wish to thank the psychologists for their willingness to devote the neces- 
sary hour or more to the painstaking task of supplying the ratings called 
for. We are also indebted to Dr. Paul Buttenwieser for assistance in preparing 
t.M summary. 



INTERPRETATIONS AND CONCLUSIONS 



455 



Ranks correlated 


For male 
judges 


For female 
judges 


For male occupational groups . 


759 


674 


For female occupational groups. . . 
For special male groups 


.428 
.504 


.663 
496 


For special female groups 


.530 


508 









These three correlations range from .66 to .76, the other five from 
.43 to .53. The average rankings by the 21 male judges correlate 
as follows with the averages of the 21 female judges: 

For male occupational groups 984 

For female occupational groups 991 

For special male groups 966 

For special female groups 978 

We can say, therefore, that neither the sex of the raters nor the 
sex of the groups being rated affects very greatly the reliability 
of the ratings. The one marked exception is in the case of the 
female occupational groups; here the average agreement among 
female judges is .66 and among the male judges .43. In general 
one can say that the agreement among judges of either sex is 
much greater than we have found for less expert judges who rated 
individuals rather than groups. Even so, the agreement repre- 
sented by an average intercorrelation around .50 is in reality very 
low measured in terms of the coefficient of alienation. The data 
suggest that even for psychologists the terms "masculinity" and 
" femininity" do not have very clear connotations. 

Our present interest, however, is in the agreement between the 
M-F scores and the average rankings of the 21 male and 21 female 
judges. 1 This is as follows for the four groups: 



Ranks correlated 


Male 
judges 


Female 
judges 


For male occupational groups 


.349 


.347 


For female occupational groups. . . 
For special male groups 


.382 
.656 


.443 
.720 


For special female groups 


.582 


.659 









1 See p. 247 for correlation between M-F scores and scale ratings of 26 occupa- 
tions of homosexual males. 



456 SEX AND PERSONALITY 

It is evident from these correlations that the female judges suc- 
ceed slightly better than do the male judges in approximating the 
rankings yielded by the test scores. On the surface, it would 
seem that the rankings of both male and female judges agree less 
closely with the test-score rankings in the case of the occupational 
groups than in the case of the special groups. This difference, 
however, is more apparent than real, and may most plausibly be 
attributed to the fact that the greater spread of masculinity 
within the special groups has facilitated the task of discriminating 
between these populations. 1 When we correct for this factor by 
adjusting the coefficients to the values that would be expected 
if the range in M-F test scores were equal for all groups, the 
ability of the judges to approximate the " correct " order becomes 
roughly the same for all four classes of populations. 

Table 82 shows for each of the four lists of populations the 
amount of agreement in the rank orders by mean M-F score and 
composite ratings of the male and female judges. The reader 
will find it interesting to examine this table rather carefully for 
evidence of constant biases. 

When we adopt the objective test scores for each group as the 
standard, and compare these means with the composite rankings 
of the judges, certain biases are apparent. Both male and female 
judges tend to underestimate the masculinity of men engaged in 
professional and intellectual pursuits, and to overestimate the 
masculinity of outdoor or manual workers. Farmers, for 
example, are ranked 5 in the male occupational list by both 
groups of judges, but IS by average M-F score. Lawyers, 
bankers, physicians, dentists, teachers (male), and architects are 
all ranked several grades more feminine by the judges than by 
the test. The idea seems to be that the place of a really mascu- 
line man is out of doors. It is unfortunate that we were not able 
to check this theory further by testing a group of cowboys. 

The most pronounced disagreement is the case of city policemen 
and firemen, who are ranked by both men and women as the most 

1 The range between the highest and lowest mean is as follows in the four 
groups: Group I, 59 points; Group II, 40 points; Group III, 112 points; Group 
IV, 79 points, 



INTERPRETATIONS AND CONCLUSIONS 



457 



TABLE 82. AGREEMENT BETWEEN MEAN M-F SCORES AND PSYCHOLOGISTS' 
RANKINGS OF SELECTED POPULATIONS 



Selected populations 


Rank 
order 
M-F test 


Composite rank 
order 


Male 
judges 


Female 
judges 


I. 

II. 
III. 

IV. 


Male occupational groups: 


3 

8 
7 
4 
12 
10 
15 
18 
1 
9 
11 
17 
19 
5 
13 
6 
14 
20 
16 

1 
4 

2 5 
11 
9 
10 
8 
6 
2 5 

7 
9 
2 
3 
1 
4 
10 

12 
6 

8 
11 

6 
4 

11 
8 
10 

13 

7 
9 
12 


7 
9 
10 
11 
14 
6 
13 
5 
IS 
4 
16 
18 
12 
19 
8 
3 
2 
20 
17 
1 

3 
4 
1 
2 
7 
8 
6 
10 
11 

9 

5 
7 
6 
3 

10 
11 

is 

8 5 

2 
6 

5 

11 
12 
9 
10 

13 
8 
7 


8 
6 
10 
11 
14 

13 

15 
4 
18 
17 
12 
20 
9 
2 

19 
16 
1 

3 

4 

6 
8 

10 
11 

9 

6 

5 

4 
1 
9 
11 

12 
8 
2 

10 

8 

4 

5 
11 
9 
6 

13 
10 
12 




Physicians 


Dentists 


Architects 


Merchants 




Farmers 


Catholic pnests in training. . . 
Professional engineers 
High-school and college teachers 
Bookkeepers and office clerks 
Journalists . ... 




General salesmen 


Machinists and mechanicians . 
Contractors and builders 




Protestant ministers 


City policemen and firemen 

Female occupational groups: 
Graduate practicing nurses 
Teachers in elementary and high schools 
Office managers and business administrators 
Bookkeepers and business clerks 
Stenographers 


Domestic servants 


Musicians 
Dressmakers and seamstresses ... 
Unselected housewives . . . 
Artists and photographers 
Hairdressers and cosmeticians 

Special male groups: 
Juvenile delinquents (typical reform-school popu- 




Unselected tenth-grade boys. . ... 
Unselected college students. . 
College football players. . 
Intellectually gifted boys (ages 16-20) 
Japanese boys (born in America of Japanese parents) 
Male homosexuals (typical male prostitutes of a large 
city) .... 
Unselected eighth-grade boys 
Unselected men in their 20's 
Unselected men in their 40's .... ... 
Unselected men in their 60's 

Special female groups: 


Female prostitutes ... . 


Gifted girls (age 16-^20) 
Women athletes and directors of physical education. 
Juvenile delinquents (inmates of typical reform-school 
for girls, age 16-20) 
Japanese girls (born in America) 
Unselected tenth-grade girls . . ... 
Unselected women college students 
Unselected eighth-grade girls . 
Women holding M.D. or Ph.D. degree . 
Unselected women in their 20's 


Unselected women in their 40's 
Unselected women in their 60's .... 



458 SEX AND PERSONALITY 

masculine occupational group, whereas they score among the 
lowest on the M-F test. In rating the latter group it appears 
that the judges failed to make allowance for the masculinizing 
effect of a male uniform and for specific occupational selection in 
terms of stature. Also they overlooked two selective factors that 
have probably entered: (1) the craving of some effeminate men 
for uniforms as a means of enhancing their feeling of masculinity, 
and (2) the avoidance of ill-paid and rather routine civil-service 
jobs by men who have in pronounced degree the masculine quali- 
ties of aggressiveness, self-confidence, and independence. 

The large overestimation of the masculinity of delinquent girls 
seems to us a relatively "good" error, as the animal psychologist 
would say, for it probably rests on the rather plausible (though 
apparently erroneous) assumption that such girls nearly always 
come into conflict with the law because of their masculine 
aggressiveness. Data presented in Chapter XIV in fact suggest 
that in some cases their troubles can be blamed to the feminine 
qualities of submissiveness and docility rather than to masculine 
independence and aggressiveness on their part. 

One of the largest disagreements between the two sets of judges 
is in the ranks assigned to women aged sixty to seventy (rank 7 
by male and 12 by female judges). There is of course an obvious 
explanation of the fact that the male not yet past middle age finds 
it hard to think of old women as feminine. The female judge 
also finds it harder than the male judge to think of old men as 
masculine, though here the difference is much less. It is safe to 
predict that the old groups would fare still worse in the ratings of 
opposite-sex judges of the late teens. 

The age-of -judge factor is also probably involved in the under- 
estimates of the masculinity of eighth-grade and tenth-grade 
boys; to the middle-aged, adolescent boys still seem immature and 
half sexed. However, to our mature female judges the eighth- 
grade and tenth-grade girls seem much more masculine than the 
test shows them to be; only the male judges are able to appreciate 
their femininity! 

In the foregoing discussion we have perhaps seemed to assume 
that when test and judges disagree the judges are always in error. 



INTERPRETATIONS AND CONCLUSIONS 459 

It can only be said that the test is right in so far as it samples 
M-F differences. To that extent it is right so far as the tested 
populations are concerned. The items composing the test may 
sample the totality of M-F differences more adequately than the 
tested groups sample the general classes to which they belong. If 
larger populations had been tested the rank orders for mean score 
might not have been quite the same. We have no wish to 
discredit the ratings our self-sacrificing friends have so kindly 
furnished us. Our little experiment will have served its purpose 
if it has directed attention to some of the types of bias which 
enter into M-F ratings and to the need of objective scores which 
are free from censure, uninfluenced by likes or dislikes, and 
immune to sex appeal. 

SUGGESTIONS FOR REVISION OF THE M-F TEST 

Revision of the M-F test should accomplish two results. (1) 
The least reliable sections should be eliminated and only those 
parts should be retained which can be made reliable enough to 
warrant profile treatment of individual scores. By combining 
Forms A and B, and by increasing the number of items, the 
desired reliability could be attained for the following subtests: 
word association, information, anger response, fear response, 
disgust response, pity response, ethical censure, and interests. 
The test of introvertive response is less hopeful. Possibly 
"interests" could be divided into (a) interests in objective things, 
occupations, and activities, and (6) interests in people and their 
relationships. (2) The test should be extended to include 
samplings of sex differences in other fields. Possibilities in this 
line include, among others, an annoyance inventory, a sense-of- 
humor test, and attitude measurement by means of the Thurstone 
technique. A test designed to measure intellectual objectivity 
and emotional bias of judgment might be worth trying, although 
it is very doubtful whether a satisfactory number of items could 
be found which would show the requisite amount of sex difference. 
More promising, perhaps, would be a memory-inventory of 
childhood preferences in the field of plays, games, and amusements. 



460 SEX AND PERSONALITY 

A simpler form of the M-F test should be devised which would 
be applicable to subjects as young as eight or ten years of age. 
The crude masculinity test used by one of us in the study of 
gifted children was applicable thus early and seems to have had 
considerable prognostic value with respect to later development. 
Probably a much better test for preadolescent children could 
now be devised. 

Until suitable revision has been accomplished we believe the 
test as it stands will serve many useful purposes both in clinical 
practice and in research. It is strongly recommended, however, 
that when the test is used for the appraisal of individual subjects 
both forms be administered. For the comparison of groups a 
single form will be found satisfactory. 

NATURE AND NURTURE AS DETERMINERS OF SEX TEMPERAMENT 

The nature-nurture problem occupies a central position in any 
theory of sex temperament. The M-F test, as we have pointed 
out in our opening chapter, rests upon no assumption as to the 
causes responsible for the individual differences it discloses. The 
aim has been to devise a test which would measure whatever 
differences may exist in the hope that this would open the way to 
an empirical estimation of the relative influence of various 
determiners. At present no one knows whether the M-F deviant 
is primarily a problem for the neurologist, biochemist, and 
endocrinologist or for the parent and educator. The question 
cannot be answered without thoroughgoing search for the 
constitutional correlates of M-F deviation. The final answer 
cannot be obtained until both endocrinology and psychometrics 
have advanced beyond their present stage, though this is no 
excuse for delaying the initiation of research on the problem at 
hand. It should be emphasized, however, that failure to find the 
sought-for correlates can never be taken as conclusive proof that 
they do not exist. On the other hand, in so far as any such 
correlates may be demonstrated the nurture hypothesis is to that 
extent weakened. 



INTERPRETATIONS AND CONCLUSIONS 461 

In a recent treatise Mead 1 has presented a mass of descriptive 
evidence favoring the extreme environmental hypothesis for the 
causes of sex differences hi personality. If her observations and 
interpretations can be taken at their face value it would not be 
easy to escape the conclusion that among human beings con- 
stitutional factors are distinctly secondary to psychological as 
determiners of the M and F temperaments. Her book is based 
upon a study of three primitive tribes in New Guinea. She 
reports that in one of these, the Arapesh, males and females both 
exhibit hi the main a single temperamental pattern, one that 
corresponds closely to the feminine pattern of present-day 
occidental cultures. A similar situation was found with the 
Mundugumors, except that in this case the single standard is 
typically masculine. The Tchambuli, on the other hand, present 
both masculine and feminine patterns, but reversed as between 
the sexes, males approximating what we should call the feminine 
hi temperament and females approximating the masculine. The 
author describes in considerable detail the cultural influences 
which she believes to be responsible for these results. 

That Mead's contribution offers impressive evidence of the 
modifiability of human temperament will be readily conceded, 
but we are by no means convinced that the case for nurture is 
as strong as a casual perusal of her book would suggest. Psychol- 
ogists who have investigated personality by means of observa- 
tional and rating techniques will inevitably question the accuracy 
of anyone's estimates of the degree of masculinity or femininity 
of behavior characterizing either an individual or a group of 
individuals. It is not to be supposed that the field anthro- 
pologist, any more than the psychologist, is immune to error 
in such estimates; indeed, because the groups under observa- 
tion by him belong to an alien culture, and because his com- 
mand of the tribal language is almost invariably limited, the 
anthropologist who attempts to rate the masculinity or femininity 
of behavior in a primitive tribal group labors under tremendous 
disadvantages, 
i Op. cit. See p. 452. 



462 SEX AND PERSONALITY 

We have shown that when subjects are rated by their teachers 
or intimate acquaintances either on general masculinity-femi- 
ninity or on specific aspects of personality related thereto, so little 
agreement is found that the pooled estimates of several inde- 
pendent judges are necessary to increase the reliability of such 
ratings to a reasonable figure. Even then we do not rule out the 
types of constant error that result from a common bias among the 
raters. When subjective methods are employed, greater or less 
bias is inevitable, however competent and honest the observer 
may be; and observers who have had a particular kind of training, 
whether in anthropology or psychology, are bound to be influenced 
by the effect of biases common to their group by the "idols of 
the den." 1 

Notwithstanding the above criticisms, the book in question is 
one of the most provocative contributions thus far made to the 
psychology of sex. Written for the general reader, it naturally 
does not contain the wealth of specific detail that would be 
necessary to enable the social scientist to judge the correctness 
of its conclusions. It does, however, present a number of obser- 
vations which clearly suggest the operation of a nature as well 
as a nurture factor. The author admits that the cultural pres- 
sures in these tribes have not succeeded in forcing acceptance by 
all individuals of the personality standards imposed. Concrete 
examples are given of individuals who have become maladjusted 

1 In the specific case at hand, it is no reflection upon Dr. Mead to call attention 
to the fact, verifiable by examination of her earlier writings, that she entered 
upon her study of sex and temperament with definite leanings toward the environ- 
mental hypothesis in the interpretation of human behavior patterns. If the 
composite verbal pictures of her three New Guinea tribes had been sketched by 
an equally competent observer of different bias, there is no way of knowing how 
they would have differed from the dramatic contrasts presented; we can only be 
certain that they would have differed. It is regrettable that an investigation 
of the type in question could not have been carried out by the joint efforts of a 
number of social scientists of widely varying experiential background, includ- 
ing, say, an anthropologist, a psychologist, a sociologist, and a psychiatrist, all 
recording their observations and making their interpretations independently. 
Unfortunately, the rapidly growing contacts of these tribes with European cul- 
tures will in a few years render such an investigation meaningless, and the 
student is left to draw from the ingenious but not infallible work of Dr. Mead 
whatever conclusions seem to him reasonable. 



INTERPRETATIONS AND CONCLUSIONS 463 

by inability to conform. The author even admits that individual 
differences within a given sex are about as great as in our own 
culture, and that the chief result of the pressures has been to 
shift the location of the distribution of differences on the M-F 
axis without appreciably diminishing its range. 

The literature of anthropology furnishes an abundance of 
cogent testimony as to the plasticity of temperament and per- 
sonality. Of the treatises bearing on this question, the above- 
mentioned book by Mead and another not less notable by 
Benedict 1 are outstanding examples. Nevertheless, valuable as 
the anthropological evidence is, it cannot be accepted as a final 
answer to the nature-nurture problem. Primitive cultures are 
rapidly becoming more rare; the interpretation of behavior offers 
many pitfalls to observers unaccustomed to think in quantitative 
terms; conclusions reached by the anthropologist's field observa- 
tions are usually not amenable to laboratory checks. For these 
and other reasons the psychologist, the physiologist, the psy- 
chiatrist, and the biochemist need not fear that their contribu- 
tions to the theory of personality are likely to be rendered 
superfluous by other approaches. 

EVIDENCE OF NURTURE INFLUENCES UPON THE M-F SCORE 

Several convergent lines of evidence have been mentioned in 
preceding chapters which point to the efficacy of nurture factors 
as at least partial determiners of an individual's M-F score. 
The latter is definitely, even though not closely, associated with 
amount of schooling, with age, with occupation, with interests, 
and with domestic milieu. Perhaps the closest association of all, 
though its degree is suggested rather than measured, is that 
between cross-parent fixation and M-F deviation toward the 
norm of the opposite sex. The data do not define the reasons 
for these or other deviations. Old men test more feminine than 
young men, but the causal factor may be either experiential or 
physiological and endocrinal. Superior culture, in the case of 
women, tends to be associated with masculinity; in the case of 
men, with femininity; but our data do not tell us whether edu- 

1 BENEDICT, RUTH, Patterns of culture, p. 291, Houghton Mifflin, 1934. 



464 SEX AND PERSONALITY 

cation causes the change, or whether it merely tends to select the 
already feminine male and the already masculine female. Simi- 
larly for occupational classification, though the selective influence 
of the occupation is more clearly evidenced than in the case of 
education. Even in instances of cross-parent fixation it is not 
easy to rule out all possible selective factors: parents may be 
more likely to foster such an attachment in that particular 
opposite-sex child who is already a deviant, or, conversely, the 
already deviant child may be the only one who is affected by the 
overcherishing parent. Accordingly, although the evidence in 
favor of a considerable nurture influence is in our opinion very 
weighty, it is by no means crucial. 

From the point of view of science progress could be made more 
rapidly if experimental and control groups of infants could be 
artificially segregated and the effects watched in them of reversing 
nurture influences. Our method for human study can, however, 
not parallel the "sacrificial" procedure of the physical sciences. 
Fortunately advance is not blocked by this condition. Com- 
parison of parent-child resemblance in sex temperament with 
resemblance between foster parent and foster child can be accom- 
plished, also comparison of resemblance between identical twins 
on the one hand and between like-sex fraternal twins on the 
other. 

Another approach would be to locate parents who belong to one 
of two extreme types with respect to the kind of influence they 
have tried to exert in shaping the sex temperaments of their 
children : (a) parents who accept the usual dichotomy as desirable 
and have endeavored to inculcate it in their sons and daughters, 
and (6) parents who adhere radically to the opposite theory and 
have done their best to counteract every influence that would 
develop in their daughters the distinctively feminine, or in their 
sons the distinctively masculine, personality. If enough parents 
of the second type could be found to permit reliable determina- 
tions, the parental influence would be measured by the M-F 
score difference separating their sons and daughters as compared 
to the difference separating the sons and daughters of the other 
parental group. In such an investigation one would of course 



INTERPRETATIONS AND CONCLUSIONS 465 

need to bear in mind that parental pressures may be largely 
nullified by subtle pressures of the larger social milieu, including 
playmates, the school, the newspaper, the theater, literature, 
industry, government, and innumerable other factors. Even so, 
we believe that a careful study of parental influences upon the sex 
temperament of offspring would be worth making. 

As to what the outcome of such investigations might be, we 
prefer not to hazard a guess. On the one hand is a respectable 
body of evidence pointing to nurture effects; on the other is the 
spectacular and ever increasing evidence from animal laboratories 
on the effects of hormone concentration upon patterns of sexual 
behavior. To assume a partisan position at the present time 
with respect to the relative influence of nature and nurture upon 
human personality is hardly warranted. 

CLINICAL AND FOLLOW-UP STUDIES OF DEVIANT-SCORING 

SUBJECTS 

It is important that we should find the reason for the low 
correlation invariably obtained between M-F scores and personal 
ratings of the subjects by their acquaintances. If the mental 
attitudes measured by the test really have so little observable 
effect upon behavior, it would hardly seem worth while to inves- 
tigate them. It may be, on the other hand, that the fault lies 
with the observers. That this is in part the explanation is indi- 
cated by the small amount of agreement between equally compe- 
tent judges in rating the same individuals. It is further indicated 
by our case studies of extreme deviants. Clinical reports upon 
subjects who test 60 or 80 M-F points above their sex mean 
present a striking contrast to the composite picture of subjects 
who test correspondingly low. The pictures are blurred by 
exceptions to the rule, but the contrast is unmistakable. We can 
only conclude that the test scores do have behavioral correlates 
but that ordinary observers lack adeptness in detecting them. 
Perhaps continued use of the test technique will help to clear up 
the confused notions which are current with regard to what 
constitutes masculinity and femininity of personality. The 



466 SEX AND PERSONALITY 

fact seems to be that most of us have not acquired the ability to 
discriminate very dearly the genuinely masculine from the 
genuinely feminine. We have not learned the art of discovering 
what is beneath the roughneck apparel of the swaggering male 
who would disguise his femininity. We are likely to overestimate 
the masculinity of the big, raw-boned, homely girl and to over- 
estimate the femininity of the girl who is petite and pretty. 
Experience has shown that it is possible to talk much more 
intelligently and intelligibly about a person's mental abilities in 
terms of test scores than in vague descriptive terms, but the need 
for objective terminology in the appraisal of masculinity and 
femininity is even greater. 

The interpretation of M-F scores can only be clarified by inten- 
sive clinical and follow-up studies of large numbers of individual 
subjects. Again the situation is paralleled in the history of 
intelligence testing. It has not been long since even the wisest 
psychologist was ignorant of the prognostic significance of, say, 
a mental-age score of 9 years earned by a six-year-old. It was 
the common opinion, also the opinion of no less an expert in child 
psychology than Stanley Hall, that such children are especially 
likely to recede to or below the average as adult life approaches. 
Not long ago psychologists as well as teachers encouraged the 
parents of a backward child to believe that the retardation of 
development would probably be made good by later acceleration. 
It was only by the testing and follow-up of thousands of subjects 
that the truth came out. The same process will have to be 
gone through before we shall know the significance of M-F 
scores. 

APPLICATIONS OF THE M-F TEST IN THE STUDY 
OF HOMOSEXUALITY 

Hardly any other phenomenon connected with personality 
presents a more challenging problem than the individual who 
finds it difficult or impossible to make heterosexual adjustments. 
From the almost complete ignorance of a large majority of people 
on this subject one would suppose that homosexuals were 



INTERPRETATIONS AND CONCLUSIONS 467 

exceedingly rare. Such is far from being the case. Estimates 
by the best informed students of the subject usually place the 
proportion of males so afflicted between 3 and 5 per cent. The 
proportion in one of the largest of American universities is 
estimated by the university medical staff 1 to be in the neighbor- 
hood of 4 per cent. Of course not all who are essentially homo- 
sexual engage in overt homosexual practices, but the proportion 
who do, at least in urban communities, must be very considerable. 
A typical American city of a million population, including pos- 
sibly two hundred thousand adult males, may support several 
hundred male prostitutes who must on any reasonable estimate 
cater to the wants of several thousand men. It is probably safe 
to say that in our western cultures an average of one man in 
30 is strongly enough inclined to homosexuality to find hetero- 
sexual relationships difficult. Much less is known regarding the 
number of homosexuals among females. Although intrasex 
"crushes" are more common among females than among males, 
it is probable that the number of true female inverts is relatively 
small. The female homosexual, for obvious reasons, finds 
heterosexual adjustment less impossible. 

Of course the incidence of an abnormality has no bearing on its 
merits as a problem for scientific investigation. It has been 
alluded to in this connection only to emphasize the importance 
of homosexuality from a social and practical point of view. 
Any recognizable physical disease which afflicted so many people 
and caused so much acute unhappiness would attract hundreds 
of scientific workers and huge research funds. Homosexuality, 
on the other hand, is at present hardly mentionable; its victims 
(if males) are hounded as criminals, and the problem attracts 
from scientists but a small fraction of the attention it deserves. 

We are hopeful that the M-F test will prove a useful tool in 
this field of investigation. It does not measure homosexuality, 
as that term is commonly used, but it does measure, roughly, 
degree of inversion of the sex temperament, and it is probably 
from inverts in this sense that homosexuals are chiefly recruited. 
For one thing, the use of the test will help to center attention 

1 In a private communication to one of the authors. 



468 SEX AND PERSONALITY 

on the developmental aspects of the abnormality, just as intelli- 
gence tests have done in the case of mental deficiency. It is 
well known that the milder grades of mental deficiency can now 
be detected years earlier than was possible a generation ago. 
The same will in time be true of the potential homosexual. 
Early identification of the latter deviant is particularly to be 
desired, because we have so much reason to believe that defects 
of personality can be compensated for and to some extent 
corrected. 

Tests of a few thousand subjects, and their follow-up into adult 
life, will ultimately tell us whether M-F scores one or two stand- 
ard deviations above or below the sex mean at this early age have 
prognostic value with respect to later social and sexual adjust- 
ment. The question involves much more than merely finding 
out whether and how commonly feminine-scoring boys and 
masculine-scoring girls develop into homosexuals. It is probable 
that many who do not so develop will experience other difficulties 
of adjustment as a direct result of their deviation. One would 
like to know whether fewer of them marry, and whether a larger 
proportion of these marriages are unhappy. 

Other uses of the test in the study of sexual inversion will 
readily suggest themselves. The need for more basic cooperation 
with biochemistry has already been emphasized. Research 
should also be devoted to the invert female, who has been little 
studied except by the psychoanalysts. 1 Improvements should 
be made in Dr. Kelly's tentative scale for measuring sexual 
inversion in males, and another "I" scale arranged for use with 
females. A single "I" scale might be devised which would be 
applicable to both sexes, though it remains to be seen whether 
this is feasible. Research is needed to clarify the relationship 
between "I" scores and M-F scores. Our data indicate that a 
man may have a very feminine M-F score but at the same time a 
low "invert" score. The value of both the M-F score and the 
"I" score should be investigated with a view to finding out what 

1 A recent study from the Laboratory of Dr. Carney Landis should be men- 
tioned: STRAKOSCH, FRANCES M., Factors in the sex life of seven hundred psycho- 
pathic women, State Hospital Press, Utica, N.Y., 1934. 



INTERPRETATIONS AND CONCLUSIONS 469 

assistance they can render in the psychological differentiation of 
active and passive types of homosexuals. 

OTHER USES or THE M-F TECHNIQUE 

When more reliable subtests are available the profile studies 
thus made possible may yield important data for clinical use. 
From the low average intercorrelations of subtests of the kind 
we have used it follows that subjects who have earned the same 
M-F total score may present radically different profiles on the 
subtests. The significance of such differences remains at present 
almost wholly unexplored. It is known that certain groups tend 
to score relatively much more masculine on some of the present 
M-F "exercises" than do other groups; what the divergences 
signify has been revealed only in part. 

Some of the data presented in Chapters VI to IX suggest that 
the M-F technique may have a contribution to make to educa- 
tional, vocational, and avocational guidance. The large differ- 
ences in mean scores of various occupational groups indicate that 
such occupations exert a selective influence. Engineers, for 
example, tend to score highly masculine, artists, theologians, and 
musicians to score highly feminine. Significant relationships 
have been found between M-F score and the scholastic grades of 
college students, and between score and avocational interests of 
adult populations of both sexes. We have not investigated the 
correlation, if any, between score and success within a given 
occupation. It would be interesting to know whether, for 
example, feminine-scoring engineers or masculine-scoring min- 
isters, artists, and musicians have less than average chance to 
attain success in their professional work. 

In Chapter IV it was shown that scores on a substitute M-F 
test yielded certain small correlations with indices of marital 
happiness. For example, happily married women tested more 
feminine than women who were unhappily married or divorced, 
divorced men more feminine than men who were unhappily 
married. The differences were in the neighborhood of three times 
their standard error. At the same time no relationship was found 
between husband-wife resemblance in M-F score and the index 



470 SEX AND PERSONALITY 

of their marital happiness. The experiment should be repeated 
with the use of the present M-F test instead of the substitute 
formerly employed. 

Thus far no use has been made of the M-F test with subjects 
who were suffering from mental disorders. It would not be 
surprising if reliable differences were found in the scores of the 
cyclothymic and schizophrenic types or if other psychoneuroses 
were found to yield characteristic patterns. 

The list of suggestions we have made for further M-F research 
is far from exhaustive and it is impossible to predict which 
fields of those that confront us will yield the most fruitful 
harvests. We shall be satisfied if this pioneer attempt to apply 
psychometric methods in the study of sex temperaments stimu- 
lates others to test the soundness of our approach. We hold no 
brief for finality in the tentative interpretations and conclusions 
to which we have been led. 



APPENDIX I 

NORMS FOR TOTAL SCORE 

TABLE 83. PERCENTILE NORMS BASED ON HIGH-SCHOOL JUNIORS 



Form A 


Form B 


Males 


Females 


Males 


Females 


Per- 
centile 


Raw 


Per- 
centile 


Raw 


Per- 
centile 


Raw 


Per- 
centile 


Raw 




scores 




scores 




scores 




scores 


scores 




scores 




scores 




scores 




100 


+ 190 


100 


+ 40 


100 


+ 185 


100 


+110 


95 


+ 161 


95 


- 12 


95 


+ 156 


95 


+ 14 


90 


+ 142 


90 


- 27 


90 


+ 144 


90 


- 1 


85 


+ 132 


85 


- 38 


85 


+ 134 


85 


- 20 


80 


+ 124 


80 


- 45 


80 


+125 


80 


- 31 


75 


+ 117 


75 


- 51 


75 


+116 


75 


- 40 


70 


+ 100 


70 


- 56 


70 


+ 105 


70 


- 47 


65 


+ 92 


65 


- 61 


65 


+ 95 


65 


- 53 


60 


+ 84 


60 


- 67 


60 


+ 89 


60 


- 62 


55 


+ 77 


55 


- 75 


55 


+ 83 


55 


- 69 


50 


+ 69 


50 


- 87 


50 


+ 79 


50 


- 75 


45 


+ 60 


45 


- 97 


45 


+ 74 


45 


- 81 


40 


+ 53 


40 


-102 


40 


+ 70 


40 


- 87 


35 


+ 48 


35 


-107 


35 


+ 68 


35 


- 92 


30 


+ 44 


30 


-115 


30 


+ 52 


30 


- 98 


25 


+ 37 


25 


-123 


25 


+ 43 


25 


-108 


20 


+ 29 


20 


-129 


20 


+ 34 


20 


-118 


15 


+ 20 


15 


-136 


15 


+ 27 


15 


-127 


10 


+ 11 


10 


-143 


10 


+ 20 


10 


-140 


5 


__ 2 


5 


-153 


5 


+ 5 


5 


-156 





- 90 





-184 





- 84 





-204 



471 



472 



SEX AND PERSONALITY 



TABLE 84. PERCENTILE NORMS BASED ON COLLEGE SOPHOMORES, AVERAGE OF 

FORMS A AND B 



Males 


Females 


Percentile scores 


Raw scores 


Percentile scores 


Raw scores 


100 


-f200 


100 


+ 41 


97.5 


+ 158 


97.5 


+ 16 


95 


+ 139 


95 


+ 8 


90 


+ 125 


90 


- 4 


85 


+ 114 


85 


- 13 


80 


+106 


80 


- 22 


75 


+ 99 


75 


- 31 


70 


+ 75 


70 


- 40 


65 


+ 90 


65 


- 46 


60 


+ 86 


60 


- 53 


55 


+ 82 


55 


- 59 


50 


+ 77 


50 


- 65 


45 


+ 72 


45 


- 71 


40 


+ 66 


40 


- 76 


35 


+ 59 


35 


- 81 


30 


+ 51 


30 


- 85 


25 


+ 41 


25 


- 90 


20 


+ 32 


20 


- 95 


15 


+ 21 


15 


-100 


10 


+ 7 


10 


-107 


5 


- 17 


5 


-117 


2 5 


- 37 


2 5 


-124 





- 70 





-140 



APPENDIX I 473 

TABLE 85. STANDARD SCORES BASED ON HIGH-SCHOOL JUNIORS 



Males 


Females 


Raw 


Standard scores 


Raw 


Standard scores 


scores 


Form A 


FormB 


scores 


Form A 


Form B 


+200 


+2.45 


+2.43 


+ 60 


+3.20 


+2.67 


+190 


+2.26 


+2.32 


+ 50 


+2.98 


+2.48 


+ 180 


+2.06 


+2.11 


+ 40 


+2.76 


+2.28 


+170 


+ 1.87 


+ 1.90 


+ 30 


+2.54 


+2.08 


+160 


+ 1.68 


+1.69 


+ 20 


+2.32 


+ 1.88 


+ 150 


+ 1.49 


+ 1.48 


+ 10 


+2.10 


+ 1 68 


+140 


+ 1.29 


+ 1.27 





+ 1.87 


+ 1.48 


+130 


+ 1.10 


+ 1.06 


- 10 


+ 1.65 


+ 1.28 


+120 


+ .91 


+ .84 


- 20 


+ 1.43 


+ 1.09 


+110 


+ .72 


+ .63 


- 30 


+ 1.21 


+ .89 


+ 100 


+ .53 


+ .42 


- 40 


+ .99 


+ .69 


+ 90 


+ .33 


+ .21 


- 50 


+ .77 


+ .49 


+ 80 


+ .14 


.00 


- 60 


+ .55 


+ .29 


+ 70 


- .05 


- .21 


- 70 


+ .33 


+ .09 


+ 60 


- .24 


- .42 


- 80 


+ .11 


- .09 


+ 50 


44 


- .63 


- 90 


- .11 


- .29 


+ 40 


- .63 


- .85 


-100 


- .34 


- .41 


+ 30 


- .82 


-1.06 


-110 


- .56 


- .69 


+ 20 


-1.01 


-1.27 


-120 


- 78 


- 89 


+ 10 


-1.21 


-1.48 


-130 


-1.00 


-1.09 





-1.40 


-1 69 


-140 


-1.22 


-1.28 


- 10 


-1.59 


-1.90 


-150 


-1.44 


-1 48 


- 20 


-1.78 


-2 11 


-160 


-1.66 


-1 68 


- 30 


-1.97 


-2.32 


-170 


-1.88 


-1.88 


- 40 


-2.17 


-2.53 


-180 


-2.10 


-2.08 


- 50 


-2.36 


-2.75 


-190 


-2.32 


-2.28 


- 60 


-2.55 


-2.96 


-200 


-2.55 


-2.48 


- 70 


-2.74 


-3 16 









APPENDIX H 

NORMS FOR EXERCISES 1 TO 7 

TABLE 86. PERCENTILE NORMS FOR THE SEPARATE EXERCISES OF FORM A 
BASED ON HIGH-SCHOOL JUNIORS 



Percentile 


Exercise 1 


Exercise 2 


Exercise 3 


Exercise 4 


scores 




















Males 


Females 


Males 


Females 


Males 


Females 


Males 


Females 


95 


+ 15 


- 1 


+2 


+ 1 


+ 18 


+ 1 


+60 


+28 


90 


+ 11 


- 2 


+ 1 


+ 1 


+ 17 


- 2 


+53 


+22 


80 


+ 6 


- 7 


+ 1 





+ 15 


- 4 


+35 


+ 5 


70 


+ 4 


- 9 








+ 12 


- 7 


+27 


- 6 


60 





-12 





-1 


+ 9 


- 9 


+21 


-11 


50 


- 1 


-13 





-1 


+ 7 


-11 


+ 15 


-14 


40 


3 


-16 


-1 


-1 


+ 5 


-13 


+ 11 


-21 


30 


- 5 


-18 


-1 


-2 


+ 3 


-17 


+ 5 


-28 


20 


- 6 


-21 


-1 


-2 


+ 1 


-20 


- 7 


-34 


10 


-10 


-22 


-2 


-3 


- 3 


-25 


-20 


-44 


5 


-14 


-24 


-2 


-3 


- 7 


-28 


-25 


-48 



Per- 


Exercise 5 


Exercise 6 


Exercise 7 


A;!- 








cenuie 
scores 


Males 


Females 


Males 


Females 


Males 


Females 


95 


+99 


+ 18 


+ 14 


+ 8 


+4 


+2 


90 


+92 


- 4 


+ 12 


+ 7 


+3 


+ 1 


80 


+79 


-10 


+ 10 


+ 1 


+2 





70 


+ 70 


-22 


+ 8 


- 1 


+ 1 





60 


+62 


-30 


+ 6 


- 3 


+ 1 


-1 


50 


+54 


-38 


+ 3 


- 5 





-2 


40 


+46 


-47 





- 7 





-2 


30 


+37 


-55 


- 3 


- 9 


-1 


-3 


20 


+19 


-61 


- 6 


-11 


-1 


-3 


10 


+ 2 


-70 


-10 


-14 


-2 


-4 


5 


- 4 


-78 


-13 


-20 


_* 


-4 



474 



APPENDIX II 



475 



TABLE 87. PERCENTILE NORMS FOR THE SEPARATE EXERCISES OF FORM B 
BASED ON HIGH-SCHOOL JUNIORS 



Percentile 


Exercise 1 


Exercise 2 


Exercise 3 


Exercise 4 


scores 




















Males 


Females 


Males 


Females 


Males 


Females 


Males 


Females 


95 


+ 13 





+3 


+ 1 


+ 16 


+ 4 


+54 


+32 


90 


+ 10 


- 4 


+2 





+ 14 


+ 2 


+48 


+26 


80 


+ 7 


- 8 


+ 1 





+ 10 


- 2 


+35 


+ 14 


70 


+ 4 


-11 


+ 1 


__ i 


+ 8 


- 4 


+27 


+ 1 


60 


+ 3 


-13 


+ 1 


-1 


+ 6 


- 5 


+21 


- 7 


SO 


+ 1 


-15 





-1 


+ 5 


- 8 


+ 14 


-14 


40 





-16 





-1 


+ 3 


-10 


+ 8 


-17 


30 


- 3 


-18 





-2 


+ 2 


-12 





-24 


20 


- 6 


-20 


-1 


-2 


- 1 


-16 


- 4 


-34 


10 


- 8 


-23 


-1 


-3 


- 7 


-21 


-15 


-44 


5 


-10 


-25 


-1 


-3 


-10 


-25 


-24 


-50 



Per- 


Exercise 5 


Exercise 6 


Exercise 7 


scores 


Males 


Females 


Males 


Females 


Males 


Females 


95 


+ 114 


+ 19 


+ 16 


+ 7 


+4 


+ 1 


90 


+ 102 


+ 14 


+ 12 


+ 4 


+3 





80 


+ 89 


- 5 


+ 7 


+ 2 


+ 1 


-1 


70 


+ 79 


-16 


+ 5 


- 1 





-1 


60 


+ 70 


-25 


+ 3 


3 





-2 


50 


+ 64 


-33 


+ 2 


- 5 


-1 


-2 


40 


+ 55 


-39 





- 7 


-1 


-3 


30 


+ 44 


-46 


- 1 


-10 


-2 


-3 


20 


+ 29 


-53 


- 4 


-13 


-2 


-4 


10 


+ 22 


-72 


-11 


-16 


-3 


-5 


5 


+ 13 


-78 


-12 


-19 


-3 


-6 



476 



SEX AND PERSONALITY 



TABLE 88. PERCENTILE NORMS FOR THE SEPARATE EXERCISES, BASED ON 
COLLEGE SOPHOMORES. AVERAGE OF FORMS A AND B 



Percentile 


Exercise 1 


Exercise 2 


Exercise 3 


Exercise 4 


scores 




















Males 


Females 


Males 


Females 


Males 


Females 


Males 


Females 


95 


4-12 


+ 1 


+2 


+ 1 


+ 18 


+ 5 


+58 


+40 


90 


+ 10 


- 2 


+ 1 





+ 15 


+ 2 


+49 


+33 


80 


+ 7 


- 5 


+ 1 





+ 13 


- 1 


+40 


+25 


70 


+ 4 


- 8 


+ 1 


-1 


+ 11 


- 3 


+34 


+ 19 


60 


+ 2 


-11 





-1 


+ 9 


- 5 


+29 


+ 13 


50 





-14 





-1 


+ 7 


- 7 


+25 


+ 6 


40 


- 2 


-16 





-1 


+ 6 


- 8 


+20 





30 


- 4 


-18 


-1 


-2 


+ 4 


-11 


+ 16 


- 6 


20 


- 7 


-20 


-1 


-2 


+ 2 


-13 


+ 10 


-12 


10 


-10 


-22 


-1 


-3 


- 2 


-17 


+ 1 


-20 


5 


-13 


-23 


-2 


-3 


- 5 


-20 


g 


-26 



Per- 


Exercise 5 


Exercise 6 


Exercise 7 


scores 


Males 


Females 


Males 


Females 


Males 


Females 


95 


+87 


+ 3 


+ 12 


+ 9 


+3 


+ 1 


90 


+76 


- 6 


+ 10 


+ 7 


+2 





80 


+63 


-20 


+ 7 


+ 4 


+ 1 


-1 


70 


+53 


-31 


+ 6 


+ 2 





-2 


60 


+43 


-40 


+ 4 





-1 


-2 


50 


+35 


-47 


+ 3 


- 1 


-1 


-3 


40 


+26 


-53 


+ 1 


- 3 


-2 


-4 


30 


+ 16 


-60 


- 1 


- 5 


-3 


-4 


20 


+ 3 


-67 


- 3 


- 7 


-3 


-5 


10 


-15 


-76 


- 5 


-10 


-4 


-6 


5 


-32 


-83 


- 7 


-13 


-6 


-6 



APPENDIX II 



477 



TABLE 89. STANDARD SCORES FOR MALES, BASED ON GENERAL ADULT 
POPULATION OF 552 CASES l 



Standard 
score 


Ex. 1 


Ex. 2 


Ex. 3 


Ex. 4 


Ex.5 


Ex.6 


Ex.7 


Total 


+3.00 


+20 


+4 


+27 


+86 


+ 127 


+22 


+5 


+ 197 


+2.75 


+ 18 


+3 


+25 


+80 


+ 118 


+20 


+5 


+ 184 


+2.50 


+ 16 


+3 


+22 


+74 


+ 110 


+ 18 


+4 


+ 171 


+2.25 


+ 14 


+3 


+20 


+69 


+ 101 


+ 16 


+4 


+ 157 


+2.00 


+ 12 


+2 


+ 18 


+63 


+ 92 


+ 14 


+3 


+ 144 


+1.75 


+ 11 


+2 


+ 16 


+57 


+ 84 


+12